{"1024-x-1024/50-steps/bedrock/amazon.nova-canvas-v1:0":{"mode":"image_generation","base_model":"nova-canvas","max_input_tokens":2600,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2600},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_nova_canvas_image_edit":true},"1024-x-1024/50-steps/stability.stable-diffusion-xl-v1":{"mode":"image_generation","base_model":"stable-diffusion-xl-v1","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1024/dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1024/max-steps/stability.stable-diffusion-xl-v1":{"mode":"image_generation","base_model":"stable-diffusion-xl-v1","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"256-x-256/dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"512-x-512/50-steps/stability.stable-diffusion-xl-v0":{"mode":"image_generation","base_model":"stable-diffusion-xl-v0","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"512-x-512/dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"512-x-512/max-steps/stability.stable-diffusion-xl-v0":{"mode":"image_generation","base_model":"stable-diffusion-xl-v0","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ai21.j2-mid-v1":{"mode":"chat","base_model":"j2-mid-v1","max_input_tokens":8191,"max_output_tokens":8191,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ai21.j2-ultra-v1":{"mode":"chat","base_model":"j2-ultra-v1","max_input_tokens":8191,"max_output_tokens":8191,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ai21.jamba-1-5-large-v1:0":{"mode":"chat","base_model":"jamba-1-5-large","deprecation_date":"2026-11-26","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-ai21-labs-jamba-1-5-large.md","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":false,"supports_function_calling":true,"supports_tool_choice":true,"supported_regions":["us-east-1"],"is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ai21.jamba-1-5-mini-v1:0":{"mode":"chat","base_model":"jamba-1-5-mini","deprecation_date":"2026-11-26","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-ai21-labs-jamba-1-5-mini.md","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1"],"is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ai21.jamba-instruct-v1:0":{"mode":"chat","base_model":"jamba-instruct","max_input_tokens":70000,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","supports_system_messages":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","metadata":{"notes":"DALL-E 2 via AI/ML API - Reliable text-to-image generation"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","metadata":{"notes":"DALL-E 3 via AI/ML API - High-quality text-to-image generation"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux-pro":{"mode":"image_generation","base_model":"flux-pro","metadata":{"notes":"Flux Dev - Development version optimized for experimentation"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux-pro/v1.1":{"mode":"image_generation","base_model":"flux-pro/v1.1","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux-pro/v1.1-ultra":{"mode":"image_generation","base_model":"flux-pro/v1.1-ultra","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux-realism":{"mode":"image_generation","base_model":"flux-realism","metadata":{"notes":"Flux Pro - Professional-grade image generation model"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux/dev":{"mode":"image_generation","base_model":"flux/dev","metadata":{"notes":"Flux Dev - Development version optimized for experimentation"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux/kontext-max/text-to-image":{"mode":"image_generation","base_model":"flux/kontext-max/text-to-image","metadata":{"notes":"Flux Pro v1.1 - Enhanced version with improved capabilities and 6x faster inference speed"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux/kontext-pro/text-to-image":{"mode":"image_generation","base_model":"flux/kontext-pro/text-to-image","metadata":{"notes":"Flux Pro v1.1 - Enhanced version with improved capabilities and 6x faster inference speed"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/flux/schnell":{"mode":"image_generation","base_model":"flux/schnell","metadata":{"notes":"Flux Schnell - Fast generation model optimized for speed"},"source":"https://docs.aimlapi.com/","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/google/imagen-4.0-ultra-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-ultra-generate","metadata":{"notes":"Imagen 4.0 Ultra Generate API - Photorealistic image generation with precise text rendering"},"source":"https://docs.aimlapi.com/api-references/image-models/google/imagen-4-ultra-generate","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aiml/google/nano-banana-pro":{"mode":"image_generation","base_model":"nano-banana-pro","metadata":{"notes":"Gemini 3 Pro Image (Nano Banana Pro) - Advanced text-to-image generation with reasoning and 4K resolution support"},"source":"https://docs.aimlapi.com/api-references/image-models/google/gemini-3-pro-image-preview","supported_endpoints":["/v1/images/generations"],"provider":"aiml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.nova-canvas-v1:0":{"mode":"image_generation","base_model":"nova-canvas","max_input_tokens":2600,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2600},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_nova_canvas_image_edit":true},"us.writer.palmyra-x4-v1:0":{"mode":"chat","base_model":"palmyra-x4","deprecation_date":"2027-03-09","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x4.md","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.writer.palmyra-x5-v1:0":{"mode":"chat","base_model":"palmyra-x5","deprecation_date":"2027-03-21","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x5.md","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"writer.palmyra-x4-v1:0":{"mode":"chat","base_model":"palmyra-x4","deprecation_date":"2027-03-09","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x4.md","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"writer.palmyra-x5-v1:0":{"mode":"chat","base_model":"palmyra-x5","deprecation_date":"2027-03-21","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x5.md","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":10000,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.nova-2-pro-preview-20251202-v1:0":{"mode":"chat","base_model":"nova-2-pro","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","max_input_tokens":1000000,"max_output_tokens":64000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.amazon.nova-2-pro-preview-20251202-v1:0":{"mode":"chat","base_model":"nova-2-pro","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://aws.amazon.com/nova/pricing/","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","max_input_tokens":1000000,"max_output_tokens":64000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.amazon.nova-2-pro-preview-20251202-v1:0":{"mode":"chat","base_model":"nova-2-pro","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://aws.amazon.com/nova/pricing/","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-2-pro-preview-20251202-v1:0":{"mode":"chat","base_model":"nova-2-pro","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.nova-2-multimodal-embeddings-v1:0":{"mode":"embedding","base_model":"nova-2-multimodal-embeddings","max_input_tokens":8172,"output_vector_size":3072,"source":"https://us-east-1.console.aws.amazon.com/bedrock/home?region=us-east-1#/model-catalog/serverless/amazon.nova-2-multimodal-embeddings-v1:0","supports_embedding_image_input":true,"supports_image_input":true,"supports_video_input":true,"supports_audio_input":true,"provider":"bedrock","max_tokens":8172},"amazon.nova-micro-v1:0":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":10000,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":10000,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"amazon.rerank-v1:0":{"mode":"rerank","base_model":"rerank","max_input_tokens":32000,"max_output_tokens":32000,"provider":"bedrock","max_document_chunks_per_query":100,"max_query_tokens":32000,"max_tokens":32000,"max_tokens_per_document_chunk":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.titan-embed-image-v1":{"mode":"embedding","base_model":"titan-embed-image-v1","max_input_tokens":128,"metadata":{"notes":"'supports_image_input' is a deprecated field. Use 'supports_embedding_image_input' instead."},"output_vector_size":1024,"source":"https://us-east-1.console.aws.amazon.com/bedrock/home?region=us-east-1#/providers?model=amazon.titan-image-generator-v1","supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":128},"amazon.titan-embed-text-v1":{"mode":"embedding","base_model":"titan-embed-text-v1","max_input_tokens":8192,"output_vector_size":1536,"provider":"bedrock","max_tokens":8192},"amazon.titan-embed-text-v2:0":{"mode":"embedding","base_model":"titan-embed-text","max_input_tokens":8192,"output_vector_size":1024,"provider_specific_entry":{"bedrock_invocation_schema":"titan_v2"},"provider":"bedrock","max_tokens":8192},"amazon.titan-image-generator-v1":{"mode":"image_generation","base_model":"titan-image-generator-v1","provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.titan-image-generator-v2":{"mode":"image_generation","base_model":"titan-image-generator-v2","deprecation_date":"2026-06-30","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-titan-image-generator-g1-v2.html","provider":"bedrock","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.titan-image-generator-v2:0":{"mode":"image_generation","base_model":"titan-image-generator","deprecation_date":"2026-06-30","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-titan-image-generator-g1-v2.html","provider":"bedrock","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"twelvelabs.marengo-embed-2-7-v1:0":{"mode":"embedding","base_model":"marengo-embed-2-7","deprecation_date":"2026-11-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":77,"output_vector_size":1024,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":77},"us.twelvelabs.marengo-embed-2-7-v1:0":{"mode":"embedding","base_model":"marengo-embed-2-7","deprecation_date":"2026-11-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":77,"output_vector_size":1024,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":77},"eu.twelvelabs.marengo-embed-2-7-v1:0":{"mode":"embedding","base_model":"marengo-embed-2-7","deprecation_date":"2026-11-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":77,"output_vector_size":1024,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":77},"twelvelabs.pegasus-1-2-v1:0":{"mode":"chat","base_model":"pegasus-1-2","deprecation_date":"2027-03-22","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.md","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.twelvelabs.pegasus-1-2-v1:0":{"mode":"chat","base_model":"pegasus-1-2","deprecation_date":"2027-03-22","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.md","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.twelvelabs.pegasus-1-2-v1:0":{"mode":"chat","base_model":"pegasus-1-2","deprecation_date":"2027-03-22","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.md","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon.titan-text-express-v1":{"mode":"chat","base_model":"titan-text-express-v1","max_input_tokens":42000,"max_output_tokens":8000,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8000,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"amazon.titan-text-lite-v1":{"mode":"chat","base_model":"titan-text-lite-v1","max_input_tokens":42000,"max_output_tokens":4000,"provider":"bedrock","max_tokens":4000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"amazon.titan-text-premier-v1:0":{"mode":"chat","base_model":"titan-text-premier","max_input_tokens":42000,"max_output_tokens":32000,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32000,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"anthropic.claude-3-5-haiku-20241022-v1:0":{"mode":"chat","base_model":"claude-3-5-haiku","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"prompt_cache_min_tokens":2048,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_web_search":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-haiku-4-5@20251001":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_web_search":true},"anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":1000000,"max_output_tokens":4096,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_cache_point":false,"is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-5-sonnet-20241022-v2:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":1000000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-7-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-opus-20240229-v1:0":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-3-sonnet-20240229-v1:0":{"mode":"chat","base_model":"claude-3-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-opus-4-1-20250805-v1:0":{"mode":"chat","base_model":"claude-opus-4-1","deprecation_date":"2027-01-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"token_count_model":{"model":"anthropic.claude-opus-4-1-20250805-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2"],"is_deprecated":true,"max_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-opus-4-20250514-v1:0":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","max_tokens":32000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-opus-4-5-20251101-v1:0":{"mode":"chat","base_model":"claude-opus-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"high","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-5-20251101-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-opus-4-6-v1":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-opus-4-6-v1":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"us.anthropic.claude-opus-4-6-v1":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-opus-4-6-v1":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"au.anthropic.claude-opus-4-6-v1":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-14","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"anthropic.claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"bedrock_converse_supports_strict_tools":false,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-20250514-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1"],"max_tokens":64000,"is_deprecated":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_assistant_prefill":true,"supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_assistant_prefill":true,"supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anyscale/HuggingFaceH4/zephyr-7b-beta":{"mode":"chat","base_model":"zephyr-7b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"anyscale","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/codellama/CodeLlama-34b-Instruct-hf":{"mode":"chat","base_model":"codellama-34b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"anyscale","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/codellama/CodeLlama-70b-Instruct-hf":{"mode":"chat","base_model":"codellama-70b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/codellama-CodeLlama-70b-Instruct-hf","provider":"anyscale","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/google/gemma-7b-it":{"mode":"chat","base_model":"gemma-7b-it","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/google-gemma-7b-it","provider":"anyscale","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anyscale/meta-llama/Llama-2-13b-chat-hf":{"mode":"chat","base_model":"llama-2-13b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"anyscale","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/meta-llama/Llama-2-70b-chat-hf":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"anyscale","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/meta-llama/Llama-2-7b-chat-hf":{"mode":"chat","base_model":"llama-2-7b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"anyscale","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/meta-llama/Meta-Llama-3-70B-Instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/meta-llama-Meta-Llama-3-70B-Instruct","provider":"anyscale","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/meta-llama/Meta-Llama-3-8B-Instruct":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/meta-llama-Meta-Llama-3-8B-Instruct","provider":"anyscale","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"anyscale/mistralai/Mistral-7B-Instruct-v0.1":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mistral-7B-Instruct-v0.1","supports_function_calling":true,"provider":"anyscale","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/mistralai/Mixtral-8x22B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x22b-instruct","max_input_tokens":65536,"max_output_tokens":65536,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mixtral-8x22B-Instruct-v0.1","supports_function_calling":true,"provider":"anyscale","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anyscale/mistralai/Mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mixtral-8x7B-Instruct-v0.1","supports_function_calling":true,"provider":"anyscale","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.amazon.nova-micro-v1:0":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":10000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.anthropic.claude-3-5-sonnet-20241022-v2:0":{"mode":"chat","base_model":"claude-3-5-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_cache_point":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-3-sonnet-20240229-v1:0":{"mode":"chat","base_model":"claude-3-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"apac.anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-14","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"apac.anthropic.claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"bedrock_converse_supports_strict_tools":false,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-20250514-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1"],"max_tokens":64000,"is_deprecated":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"assemblyai/best":{"mode":"audio_transcription","base_model":"best","provider":"assemblyai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"assemblyai/nano":{"mode":"audio_transcription","base_model":"nano","provider":"assemblyai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"au.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"au.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"azure/ada":{"mode":"embedding","base_model":"ada","deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":8191,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","max_tokens":8191},"azure/codex-mini":{"mode":"responses","base_model":"codex-mini","deprecation_date":"2026-11-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/command-r-plus":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true},"azure/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2027-03-16","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","successor_model":"azure_ai/claude-haiku-5-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","provider":"azure","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/claude-opus-4-5":{"mode":"chat","base_model":"claude-opus-4-5","deprecation_date":"2027-03-16","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","successor_model":"azure_ai/claude-opus-5-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"supports_web_search":true,"prompt_cache_min_tokens":4096,"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","provider":"azure","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/claude-opus-4-1":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-11-30","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/claude-sonnet-5-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"prompt_cache_min_tokens":1024,"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","provider":"azure","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/computer-use-preview":{"mode":"chat","base_model":"computer-use","max_input_tokens":8192,"max_output_tokens":1024,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/container":{"mode":"chat","base_model":"container","source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","deprecation_date":"2027-09-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/model_router":{"mode":"chat","base_model":"model-router","deprecation_date":"2027-05-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":200000,"max_output_tokens":32768,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-services/","comment":"Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/<deployment-name> where deployment-name is your Azure deployment (e.g., azure-model-router)","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-03-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supports_prompt_caching":true,"is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-4o-mini-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-realtime","max_input_tokens":128000,"max_output_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-4o-realtime-preview-2024-10-01":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-4o-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-5-2025-08-07":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-5-mini-2025-08-07":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/gpt-5.1-chat":{"mode":"chat","base_model":"gpt-5.1-chat","deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","is_deprecated":true,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_native_streaming":true,"supports_pdf_input":true,"supports_system_messages":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-5.1-codex":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.2-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/eu/gpt-5-nano-2025-08-07":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-nano","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/o1-2024-12-17":{"mode":"chat","base_model":"o1","deprecation_date":"2026-11-19","successor_model":"azure/eu/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/o1-mini-2024-09-12":{"mode":"chat","base_model":"o1-mini","max_input_tokens":128000,"max_output_tokens":65536,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false,"provider":"azure","max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/o1-preview-2024-09-12":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/o3-mini-2025-01-31":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global-standard/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global-standard/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-03-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supports_prompt_caching":true,"is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global-standard/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_prompt_caching":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/global/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-03-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/global/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/global/gpt-5.1-chat":{"mode":"chat","base_model":"gpt-5.1-chat","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/global/gpt-5.1-codex":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/global/gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":4097,"max_output_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-3.5-turbo-0125":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2025-03-31","max_input_tokens":16384,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_system_messages":true,"is_deprecated":true},"azure/gpt-3.5-turbo-instruct-0914":{"mode":"completion","base_model":"gpt-3.5-turbo-instruct","max_input_tokens":4097,"max_output_tokens":4097,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure_text","max_tokens":4097,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4097},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-35-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":4097,"max_output_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-35-turbo-0125":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2025-05-31","max_input_tokens":16384,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"azure/gpt-35-turbo-0301":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2025-02-13","max_input_tokens":4097,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"azure/gpt-35-turbo-0613":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2025-02-13","max_input_tokens":4097,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"azure/gpt-35-turbo-1106":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2025-03-31","max_input_tokens":16384,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"azure/gpt-35-turbo-16k":{"mode":"chat","base_model":"gpt-3.5-turbo-16k","max_input_tokens":16385,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-35-turbo-16k-0613":{"mode":"chat","base_model":"gpt-3.5-turbo-16k","max_input_tokens":16385,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-35-turbo-instruct":{"mode":"completion","base_model":"gpt-3.5-turbo-instruct","max_input_tokens":4097,"max_output_tokens":4097,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure_text","max_tokens":4097,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4097},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-35-turbo-instruct-0914":{"mode":"completion","base_model":"gpt-3.5-turbo-instruct","max_input_tokens":4097,"max_output_tokens":4097,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure_text","max_tokens":4097,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4097},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4":{"mode":"chat","base_model":"gpt-4","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_system_messages":true},"azure/gpt-4-0125-preview":{"mode":"chat","base_model":"gpt-4","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4-0613":{"mode":"chat","base_model":"gpt-4","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_system_messages":true},"azure/gpt-4-1106-preview":{"mode":"chat","base_model":"gpt-4","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4-32k":{"mode":"chat","base_model":"gpt-4-32k","max_input_tokens":32768,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4-32k-0613":{"mode":"chat","base_model":"gpt-4-32k","max_input_tokens":32768,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4-turbo":{"mode":"chat","base_model":"gpt-4-turbo","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4-turbo-2024-04-09":{"mode":"chat","base_model":"gpt-4-turbo","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4-turbo-vision-preview":{"mode":"chat","base_model":"gpt-4-turbo-vision","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.1-2025-04-14":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-mini","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.1-mini-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-mini","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-nano","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.1-nano-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-nano","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4.5-preview":{"mode":"chat","base_model":"gpt-4.5","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-4o":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_web_search":true},"azure/gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-03-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-audio-2025-08-28":{"mode":"chat","base_model":"gpt-audio","deprecation_date":"2027-03-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-audio-mini-2025-10-06":{"mode":"chat","base_model":"gpt-audio-mini","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-audio-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-audio","max_input_tokens":128000,"max_output_tokens":16384,"supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-4o-mini-audio-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-audio","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-mini-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-realtime","max_input_tokens":128000,"max_output_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-realtime-2025-08-28":{"mode":"chat","base_model":"gpt-realtime","source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","deprecation_date":"2027-03-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-realtime-mini-2025-10-06":{"mode":"chat","base_model":"gpt-realtime-mini","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-mini-transcribe":{"mode":"audio_transcription","base_model":"gpt-4o-mini-transcribe","deprecation_date":"2027-06-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":16000,"max_output_tokens":2000,"supported_endpoints":["/v1/audio/transcriptions"],"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-mini-tts":{"mode":"audio_speech","base_model":"gpt-4o-mini-tts","deprecation_date":"2027-06-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text","audio"],"supported_output_modalities":["audio"],"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-realtime-preview-2024-10-01":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-transcribe":{"mode":"audio_transcription","base_model":"gpt-4o-transcribe","deprecation_date":"2026-10-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":16000,"max_output_tokens":2000,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/audio/transcriptions"],"provider":"azure","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-4o-transcribe-diarize":{"mode":"audio_transcription","base_model":"gpt-4o-transcribe","deprecation_date":"2027-04-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":16000,"max_output_tokens":2000,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/audio/transcriptions"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-2025-11-13":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_minimal_reasoning_effort":true,"provider":"azure","max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5.1-chat-2025-11-13":{"mode":"chat","base_model":"gpt-5.1-chat","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-codex-2025-11-13":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.2-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-codex-mini-2025-11-13":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-2025-08-07":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-chat":{"mode":"chat","base_model":"gpt-5-chat","deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/blog/gpt-5-in-azure-ai-foundry-the-future-of-ai-apps-and-agents-starts-here/","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_prompt_caching":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5-chat-latest":{"mode":"chat","base_model":"gpt-5-chat","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-codex":{"mode":"responses","base_model":"gpt-5-codex","deprecation_date":"2027-03-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.1-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-mini-2025-08-07":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-nano","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-nano-2025-08-07":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4-nano","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5-pro":{"mode":"responses","base_model":"gpt-5-pro","deprecation_date":"2027-04-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.2-pro","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure?pivots=azure-openai&tabs=global-standard-aoai%2Cstandard-chat-completions%2Cglobal-standard#gpt-5","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-5.1-chat":{"mode":"chat","base_model":"gpt-5.1-chat","deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","is_deprecated":true,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_native_streaming":true,"supports_pdf_input":true,"supports_system_messages":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-codex":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.2-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-codex-max":{"mode":"responses","base_model":"gpt-5.1-codex-max","deprecation_date":"2027-05-18","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","deprecation_date":"2027-06-08","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-2025-12-11":{"mode":"chat","base_model":"gpt-5.2","deprecation_date":"2027-06-08","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-chat":{"mode":"chat","base_model":"gpt-5.2-chat","deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_native_streaming":true,"supports_pdf_input":true,"supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-chat-2025-12-11":{"mode":"chat","base_model":"gpt-5.2-chat","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-codex":{"mode":"responses","base_model":"gpt-5.2-codex","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.3-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-pro":{"mode":"responses","base_model":"gpt-5.2-pro","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-5.2-pro-2025-12-11":{"mode":"responses","base_model":"gpt-5.2-pro","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","deprecation_date":"2026-10-23","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-image-1.5","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","supports_prompt_caching":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/hd/1024-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/hd/1024-x-1792/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/hd/1792-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","deprecation_date":"2027-04-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","deprecation_date":"2026-12-16","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-image-2","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","deprecation_date":"2026-12-16","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-image-2","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1024-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1024-x-1536/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/low/1536-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1024-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1024-x-1536/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/medium/1536-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1024-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1024-x-1536/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/high/1536-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-large-2402":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":32000,"supports_function_calling":true,"provider":"azure","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-large-latest":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/o1":{"mode":"chat","base_model":"o1","deprecation_date":"2026-11-19","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/o1-2024-12-17":{"mode":"chat","base_model":"o1","deprecation_date":"2026-11-19","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/o1-mini":{"mode":"chat","base_model":"o1-mini","max_input_tokens":128000,"max_output_tokens":65536,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"azure","max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o1-mini-2024-09-12":{"mode":"chat","base_model":"o1-mini","max_input_tokens":128000,"max_output_tokens":65536,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"azure","max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o1-preview":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/o1-preview-2024-09-12":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/o3":{"mode":"chat","base_model":"o3","deprecation_date":"2026-11-19","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o3-2025-04-16":{"mode":"chat","base_model":"o3","deprecation_date":"2026-04-16","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o3-deep-research":{"mode":"responses","base_model":"o3","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o3-mini":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o3-mini-2025-01-31":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o3-pro":{"mode":"responses","base_model":"o3-pro","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/o3-pro-2025-06-10":{"mode":"responses","base_model":"o3-pro","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/o4-mini":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/o4-mini-2025-04-16":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/standard/1024-x-1024/dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/standard/1024-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/standard/1024-x-1792/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/standard/1792-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/text-embedding-3-large":{"mode":"embedding","base_model":"text-embedding-3-large","deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":8191,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","max_tokens":8191},"azure/text-embedding-3-small":{"mode":"embedding","base_model":"text-embedding-3-small","deprecation_date":"2026-04-30","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":8191,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","is_deprecated":true,"max_tokens":8191},"azure/text-embedding-ada-002":{"mode":"embedding","base_model":"text-embedding-ada-002","deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":8191,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","max_tokens":8191},"azure/speech/azure-tts":{"mode":"audio_speech","base_model":"speech/azure-tts","source":"https://azure.microsoft.com/en-us/pricing/calculator/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/speech/azure-tts-hd":{"mode":"audio_speech","base_model":"speech/azure-tts-hd","source":"https://azure.microsoft.com/en-us/pricing/calculator/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/tts-1":{"mode":"audio_speech","base_model":"tts-1","deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/tts-1-hd":{"mode":"audio_speech","base_model":"tts-1-hd","deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-4.1-2025-04-14":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-4.1-mini-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-mini","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-4.1-nano-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2026-11-04","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-nano","max_input_tokens":1047576,"max_output_tokens":32768,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-03-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supports_prompt_caching":true,"is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-4o-mini-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-realtime","max_input_tokens":128000,"max_output_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-4o-realtime-preview-2024-10-01":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-4o-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","supports_prompt_caching":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-5-2025-08-07":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-5-mini-2025-08-07":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-5-nano-2025-08-07":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-nano","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/gpt-5.1-chat":{"mode":"chat","base_model":"gpt-5.1-chat","deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-chat-latest","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","is_deprecated":true,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_native_streaming":true,"supports_pdf_input":true,"supports_system_messages":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-5.1-codex":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.2-codex","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2027-05-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/us/o1-2024-12-17":{"mode":"chat","base_model":"o1","deprecation_date":"2026-11-19","successor_model":"azure/us/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/o1-mini-2024-09-12":{"mode":"chat","base_model":"o1-mini","max_input_tokens":128000,"max_output_tokens":65536,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false,"provider":"azure","max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/o1-preview-2024-09-12":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/o3-2025-04-16":{"mode":"chat","base_model":"o3","deprecation_date":"2026-04-16","successor_model":"azure/us/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/o3-mini-2025-01-31":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us/o4-mini-2025-04-16":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/whisper-1":{"mode":"audio_transcription","base_model":"whisper-1","deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://management.azure.com/subscriptions/a88257a5-3323-4ef9-95bf-5681a46c9624/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Cohere-embed-v3-english":{"mode":"embedding","base_model":"embed-v3-english","deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/Cohere-Embed-V5-Fast","max_input_tokens":512,"output_vector_size":1024,"source":"https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice","supports_embedding_image_input":true,"provider":"azure","max_tokens":512},"azure/Cohere-embed-v3-multilingual":{"mode":"embedding","base_model":"embed-v3-multilingual","deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/Cohere-Embed-V5-Fast","max_input_tokens":512,"output_vector_size":1024,"source":"https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice","supports_embedding_image_input":true,"provider":"azure","max_tokens":512},"azure/FLUX-1.1-pro":{"mode":"image_generation","base_model":"flux-1.1-pro","deprecation_date":"2027-01-05","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":5000,"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/black-forest-labs-flux-1-kontext-pro-and-flux1-1-pro-now-available-in-azure-ai-f/4434659","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/FLUX.1-Kontext-pro":{"mode":"image_generation","base_model":"flux.1-kontext-pro","deprecation_date":"2027-01-05","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":5000,"source":"https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/flux.2-pro":{"mode":"image_generation","base_model":"flux.2-pro","deprecation_date":"2027-01-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":32000,"source":"https://ai.azure.com/explore/models/flux.2-pro/version/1/registry/azureml-blackforestlabs","supported_endpoints":["/v1/images/generations"],"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Llama-3.2-11B-Vision-Instruct":{"mode":"chat","base_model":"llama-3.2-11b-vision-instruct","max_input_tokens":128000,"max_output_tokens":2048,"max_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Llama-3.2-90B-Vision-Instruct":{"mode":"chat","base_model":"llama-3.2-90b-vision-instruct","max_input_tokens":128000,"max_output_tokens":2048,"max_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","deprecation_date":"2027-01-31","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"azure","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/blog/introducing-the-llama-4-herd-in-azure-ai-foundry-and-azure-databricks/","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","deprecation_date":"2027-01-31","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":10000000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/blog/introducing-the-llama-4-herd-in-azure-ai-foundry-and-azure-databricks/","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/Meta-Llama-3-70B-Instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":2048,"max_tokens":2048,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true},"azure/Meta-Llama-3.1-405B-Instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"max_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice","supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true},"azure/Meta-Llama-3.1-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"max_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice","supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true},"azure/Meta-Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"max_tokens":2048,"source":"https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice","supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true},"azure/Phi-3-medium-128k-instruct":{"mode":"chat","base_model":"phi-3-medium-128k-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3-medium-4k-instruct":{"mode":"chat","base_model":"phi-3-medium-4k-instruct","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3-mini-128k-instruct":{"mode":"chat","base_model":"phi-3-mini-128k-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3-mini-4k-instruct":{"mode":"chat","base_model":"phi-3-mini-4k-instruct","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3-small-128k-instruct":{"mode":"chat","base_model":"phi-3-small-128k-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3-small-8k-instruct":{"mode":"chat","base_model":"phi-3-small-8k-instruct","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3.5-MoE-instruct":{"mode":"chat","base_model":"phi-3.5-moe-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3.5-mini-instruct":{"mode":"chat","base_model":"phi-3.5-mini-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":false,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-3.5-vision-instruct":{"mode":"chat","base_model":"phi-3.5-vision-instruct","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/phi-3/","supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-4":{"mode":"chat","base_model":"phi-4","deprecation_date":"2027-01-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"microsoft--phi-4","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://techcommunity.microsoft.com/blog/machinelearningblog/affordable-innovation-unveiling-the-pricing-of-phi-3-slms-on-models-as-a-service/4156495","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/Phi-4-mini-instruct":{"mode":"chat","base_model":"phi-4-mini-instruct","deprecation_date":"2027-01-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"microsoft--phi-4","max_input_tokens":131072,"max_output_tokens":4096,"source":"https://techcommunity.microsoft.com/blog/Azure-AI-Services-blog/announcing-new-phi-pricing-empowering-your-business-with-small-language-models/4395112","supports_function_calling":true,"supports_pdf_input":false,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-4-multimodal-instruct":{"mode":"chat","base_model":"phi-4-multimodal-instruct","deprecation_date":"2027-01-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"microsoft--phi-4","max_input_tokens":131072,"max_output_tokens":4096,"source":"https://techcommunity.microsoft.com/blog/Azure-AI-Services-blog/announcing-new-phi-pricing-empowering-your-business-with-small-language-models/4395112","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_vision":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-4-mini-reasoning":{"mode":"chat","base_model":"phi-4-mini","deprecation_date":"2027-01-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"microsoft--phi-4","max_input_tokens":131072,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/","supports_function_calling":true,"supports_pdf_input":false,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/Phi-4-reasoning":{"mode":"chat","base_model":"phi-4","deprecation_date":"2027-01-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"microsoft--phi-4","max_input_tokens":32768,"max_output_tokens":4096,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_reasoning":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/mistral-document-ai-2505":{"mode":"ocr","base_model":"mistral-document-ai","supported_endpoints":["/v1/ocr"],"source":"https://devblogs.microsoft.com/foundry/whats-new-in-azure-ai-foundry-august-2025/#mistral-document-ai-(ocr)-%E2%80%94-serverless-in-foundry","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/doc-intelligence/prebuilt-read":{"mode":"ocr","base_model":"doc-intelligence/prebuilt-read","supported_endpoints":["/v1/ocr"],"source":"https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/doc-intelligence/prebuilt-layout":{"mode":"ocr","base_model":"doc-intelligence/prebuilt-layout","supported_endpoints":["/v1/ocr"],"source":"https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/doc-intelligence/prebuilt-document":{"mode":"ocr","base_model":"doc-intelligence/prebuilt-document","supported_endpoints":["/v1/ocr"],"source":"https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/","provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/MAI-DS-R1":{"mode":"chat","base_model":"mai-ds-r1","deprecation_date":"2026-02-27","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/retired-models","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/","supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/cohere-rerank-v3-english":{"mode":"rerank","base_model":"rerank-v3-english","max_input_tokens":4096,"max_output_tokens":4096,"max_query_tokens":2048,"max_tokens":4096,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/cohere-rerank-v3-multilingual":{"mode":"rerank","base_model":"rerank-v3-multilingual","max_input_tokens":4096,"max_output_tokens":4096,"max_query_tokens":2048,"max_tokens":4096,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/cohere-rerank-v3.5":{"mode":"rerank","base_model":"rerank","max_input_tokens":4096,"max_output_tokens":4096,"max_query_tokens":2048,"max_tokens":4096,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/cohere-rerank-v4.0-pro":{"mode":"rerank","base_model":"rerank-v4.0-pro","deprecation_date":"2027-09-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/cohere-rerank-v5-pro","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-cohere-rerank-4-0-in-microsoft-foundry/4477076","provider":"azure","max_query_tokens":4096,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/cohere-rerank-v4.0-fast":{"mode":"rerank","base_model":"rerank-v4.0-fast","deprecation_date":"2027-09-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/cohere-rerank-v5-fast","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-cohere-rerank-4-0-in-microsoft-foundry/4477076","provider":"azure","max_query_tokens":4096,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/deepseek-v3.2":{"mode":"chat","base_model":"deepseek","deprecation_date":"2027-11-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/deepseek-v4.1-flash","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/deepseek-v3.2-speciale":{"mode":"chat","base_model":"deepseek-v3.2","deprecation_date":"2027-11-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/deepseek-v4-pro","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","deprecation_date":"2026-08-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/deepseek-v4-pro","source":"https://techcommunity.microsoft.com/blog/machinelearningblog/deepseek-r1-improved-performance-higher-limits-and-transparent-pricing/4386367","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_reasoning":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/deepseek-v3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://techcommunity.microsoft.com/blog/machinelearningblog/announcing-deepseek-v3-on-azure-ai-foundry-and-github/4390438","supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_native_effort":true},"azure/deepseek-v3-0324":{"mode":"chat","base_model":"deepseek-v3","deprecation_date":"2026-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/deepseek-v4-flash","source":"https://techcommunity.microsoft.com/blog/machinelearningblog/announcing-deepseek-v3-on-azure-ai-foundry-and-github/4390438","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/embed-v-4-0":{"mode":"embedding","base_model":"embed-v-4-0","deprecation_date":"2027-06-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/Cohere-Embed-V5-Fast","max_input_tokens":128000,"output_vector_size":3072,"source":"https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice","supported_endpoints":["/v1/embeddings"],"supported_modalities":["text","image"],"supports_embedding_image_input":true,"provider":"azure","max_tokens":128000},"azure/global/grok-3":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://devblogs.microsoft.com/foundry/announcing-grok-3-and-grok-3-mini-on-azure-ai-foundry/","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"azure/global/grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://devblogs.microsoft.com/foundry/announcing-grok-3-and-grok-3-mini-on-azure-ai-foundry/","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"azure/grok-3":{"mode":"chat","base_model":"grok-3","deprecation_date":"2026-05-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/grok-4","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","deprecation_date":"2026-05-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/grok-4-1-fast-reasoning","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/grok-4":{"mode":"chat","base_model":"grok-4","deprecation_date":"2027-01-31","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/grok-4.3","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/","supports_function_calling":true,"supports_pdf_input":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/grok-4-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-fast-non","deprecation_date":"2026-05-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/grok-4-1-fast-non-reasoning","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/grok-4-fast-reasoning":{"mode":"chat","base_model":"grok-4-fast","deprecation_date":"2026-05-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/grok-4-1-fast-reasoning","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/grok-code-fast-1":{"mode":"chat","base_model":"grok-code-fast-1","deprecation_date":"2027-01-31","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/","supports_function_calling":true,"supports_pdf_input":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/jais-30b-chat":{"mode":"chat","base_model":"jais-30b-chat","deprecation_date":"2026-01-30","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/retired-models","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://azure.microsoft.com/en-us/products/ai-services/ai-foundry/models/jais-30b-chat","supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/jamba-instruct":{"mode":"chat","base_model":"jamba-instruct","max_input_tokens":70000,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","deprecation_date":"2027-01-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/kimi-k2-5-now-in-microsoft-foundry/4492321","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_prompt_caching":true,"provider":"azure","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/ministral-3b":{"mode":"chat","base_model":"ministral-3b","deprecation_date":"2026-12-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"ministral-ministral-4b-instruct","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-large":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-large-2407":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview","supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true},"azure/mistral-large-3":{"mode":"chat","base_model":"mistral-large-3","max_input_tokens":256000,"max_output_tokens":8191,"source":"https://azure.microsoft.com/en-us/blog/introducing-mistral-large-3-in-microsoft-foundry-open-capable-and-ready-for-production-workloads/","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-medium-2505":{"mode":"chat","base_model":"mistral-medium","deprecation_date":"2026-12-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/mistral-medium-3-5","max_input_tokens":131072,"max_output_tokens":8191,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"azure","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-nemo":{"mode":"chat","base_model":"mistral-nemo","deprecation_date":"2026-01-30","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/retired-models","successor_model":"azure_ai/mistral-small-2503","max_input_tokens":131072,"max_output_tokens":4096,"source":"https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice","supports_function_calling":true,"supports_pdf_input":false,"provider":"azure","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/mistral-small":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_reasoning":true,"supports_response_schema":true},"azure/mistral-small-2503":{"mode":"chat","base_model":"mistral-small","deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"mistralai-mistral-small-4-119b-2603","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"babbage-002":{"mode":"completion","base_model":"babbage-002","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16384,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","provider":"text-completion-openai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":false,"supports_tool_choice":false,"supports_parallel_function_calling":false,"supports_reasoning":false,"supports_response_schema":false,"supports_prompt_caching":false,"supports_web_search":false,"supports_service_tier":false,"supports_assistant_prefill":false,"supports_system_messages":false},"bedrock/*/1-month-commitment/cohere.command-light-text-v14":{"mode":"chat","base_model":"command-light-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/*/1-month-commitment/cohere.command-text-v14":{"mode":"chat","base_model":"command-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/*/6-month-commitment/cohere.command-light-text-v14":{"mode":"chat","base_model":"command-light-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/*/6-month-commitment/cohere.command-text-v14":{"mode":"chat","base_model":"command-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-northeast-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/ap-south-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-south-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-southeast-3/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-southeast-3/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html","max_input_tokens":196000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-southeast-3/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ap-southeast-3/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/ca-central-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/ca-central-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-north-1/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-north-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-north-1/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/1-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/1-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/1-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/6-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/6-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/6-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-central-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/eu-west-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-2/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/eu-west-2/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-2/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-2/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-3/mistral.mistral-7b-instruct-v0:2":{"mode":"chat","base_model":"mistral-7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/eu-west-3/mistral.mistral-large-2402-v1:0":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.md","supports_function_calling":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-west-3/mistral.mixtral-8x7b-instruct-v0:1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-south-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html","max_input_tokens":196000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/eu-south-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/invoke/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":4096,"metadata":{"notes":"Anthropic via Invoke route does not currently support pdf input."},"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/sa-east-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/sa-east-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/1-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/1-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/1-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/6-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/6-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/6-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-east-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/mistral.mistral-7b-instruct-v0:2":{"mode":"chat","base_model":"mistral-7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-east-1/mistral.mistral-large-2402-v1:0":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.md","supports_function_calling":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/mistral.mixtral-8x7b-instruct-v0:1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-1/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-2/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-2/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-2/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-2/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-east-2/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/amazon.titan-embed-text-v1":{"mode":"embedding","base_model":"titan-embed-text-v1","max_input_tokens":8192,"output_vector_size":1536,"provider":"bedrock","max_tokens":8192},"bedrock/us-gov-east-1/amazon.titan-embed-text-v2:0":{"mode":"embedding","base_model":"titan-embed-text","max_input_tokens":8192,"output_vector_size":1024,"provider":"bedrock","max_tokens":8192},"bedrock/us-gov-east-1/amazon.titan-text-express-v1":{"mode":"chat","base_model":"titan-text-express-v1","max_input_tokens":42000,"max_output_tokens":8000,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/amazon.titan-text-lite-v1":{"mode":"chat","base_model":"titan-text-lite-v1","max_input_tokens":42000,"max_output_tokens":4000,"provider":"bedrock","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/amazon.titan-text-premier-v1:0":{"mode":"chat","base_model":"titan-text-premier","max_input_tokens":42000,"max_output_tokens":32000,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","deprecation_date":"2026-07-30","max_tokens":8192,"supports_cache_point":false,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-3-haiku.md","max_input_tokens":200000,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","successor_model":"bedrock/us-gov-east-1/anthropic.claude-haiku-4-5-20251001-v1:0","supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":4096,"supports_cache_point":false,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/"},"bedrock/us-gov-east-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8000,"max_output_tokens":2048,"supports_pdf_input":true,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8000,"max_output_tokens":2048,"supports_pdf_input":true,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/amazon.titan-embed-text-v1":{"mode":"embedding","base_model":"titan-embed-text-v1","max_input_tokens":8192,"output_vector_size":1536,"provider":"bedrock","max_tokens":8192},"bedrock/us-gov-west-1/amazon.titan-embed-text-v2:0":{"mode":"embedding","base_model":"titan-embed-text","max_input_tokens":8192,"output_vector_size":1024,"provider":"bedrock","max_tokens":8192},"bedrock/us-gov-west-1/amazon.titan-text-express-v1":{"mode":"chat","base_model":"titan-text-express-v1","max_input_tokens":42000,"max_output_tokens":8000,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/amazon.titan-text-lite-v1":{"mode":"chat","base_model":"titan-text-lite-v1","max_input_tokens":42000,"max_output_tokens":4000,"provider":"bedrock","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/amazon.titan-text-premier-v1:0":{"mode":"chat","base_model":"titan-text-premier","max_input_tokens":42000,"max_output_tokens":32000,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","deprecation_date":"2026-07-30","max_tokens":8192,"supports_cache_point":false,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-3-haiku.md","max_input_tokens":200000,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","successor_model":"bedrock/us-gov-west-1/anthropic.claude-haiku-4-5-20251001-v1:0","supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":4096,"supports_cache_point":false,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/"},"bedrock/us-gov-west-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-05-09","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":8000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","supports_pdf_input":true,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-05-09","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":8000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","supports_pdf_input":true,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-1/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-west-1/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/1-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/1-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/1-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/6-month-commitment/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/6-month-commitment/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/6-month-commitment/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/anthropic.claude-instant-v1":{"mode":"chat","base_model":"claude-instant-v1","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/anthropic.claude-v1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/anthropic.claude-v2:1":{"mode":"chat","base_model":"claude","max_input_tokens":100000,"max_output_tokens":8191,"supports_tool_choice":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/mistral.mistral-7b-instruct-v0:2":{"mode":"chat","base_model":"mistral-7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-west-2/mistral.mistral-large-2402-v1:0":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.md","supports_function_calling":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/mistral.mixtral-8x7b-instruct-v0:1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_function_calling":true,"supports_reasoning":true,"provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us-west-2/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock/us.anthropic.claude-3-5-haiku-20241022-v1:0":{"mode":"chat","base_model":"claude-3-5-haiku","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"prompt_cache_min_tokens":2048,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cerebras/llama-3.3-70b":{"mode":"chat","base_model":"llama-3.3-70b","deprecation_date":"2026-02-16","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"cerebras","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"cerebras/llama3.1-70b":{"mode":"chat","base_model":"llama-3.1-70b","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"cerebras","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_system_messages":true},"cerebras/llama3.1-8b":{"mode":"chat","base_model":"llama-3.1-8b","deprecation_date":"2026-05-27","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"cerebras","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"cerebras/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.cerebras.ai/blog/openai-gpt-oss-120b-runs-fastest-on-cerebras","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cerebras","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"cerebras/qwen-3-32b":{"mode":"chat","base_model":"qwen3-32b","deprecation_date":"2026-02-16","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://inference-docs.cerebras.ai/support/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"cerebras","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"cerebras/zai-glm-4.6":{"mode":"chat","base_model":"glm-4.6","deprecation_date":"2026-01-20","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://www.cerebras.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"cerebras","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"cerebras/zai-glm-4.7":{"mode":"chat","base_model":"glm-4.7","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://www.cerebras.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"cerebras","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chat-bison":{"mode":"chat","base_model":"chat-bison","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chat-bison-32k":{"mode":"chat","base_model":"chat-bison-32k","max_input_tokens":32000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chat-bison-32k@002":{"mode":"chat","base_model":"chat-bison-32k","max_input_tokens":32000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chat-bison@001":{"mode":"chat","base_model":"chat-bison","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chat-bison@002":{"mode":"chat","base_model":"chat-bison","deprecation_date":"2025-04-09","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"chatdolphin":{"mode":"chat","base_model":"chatdolphin","max_input_tokens":16384,"max_output_tokens":16384,"provider":"nlp_cloud","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chatgpt-4o-latest":{"mode":"chat","base_model":"chatgpt-4o","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-4o-transcribe-diarize":{"mode":"audio_transcription","base_model":"gpt-4o-transcribe","max_input_tokens":16000,"max_output_tokens":2000,"supported_endpoints":["/v1/audio/transcriptions"],"deprecation_date":"2027-02-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-live-transcribe","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"claude-haiku-4-5-20251001":{"mode":"chat","base_model":"claude-haiku-4-5","supports_web_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2026-10-15","is_deprecated":true,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2026-10-15","supports_web_search":true,"is_deprecated":true,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"claude-3-haiku-20240307":{"mode":"chat","base_model":"claude-3-haiku","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":264,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"deprecation_date":"2026-04-20","is_deprecated":true},"claude-4-opus-20250514":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"claude-4-sonnet-20250514":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-09-29","deprecation_source":"https://platform.claude.com/docs/en/about-claude/model-deprecations","successor_model":"claude-sonnet-5-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"claude-sonnet-4-5-20250929":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-09-29","deprecation_source":"https://platform.claude.com/docs/en/about-claude/model-deprecations","successor_model":"claude-sonnet-5-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"prompt_cache_min_tokens":1024,"source":"https://docs.anthropic.com/en/docs/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.md","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"claude-opus-4-1":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"anthropic","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":32000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"deprecation_date":"2026-08-05","is_deprecated":true,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_native_structured_output":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_web_search":true},"claude-opus-4-1-20250805":{"mode":"chat","base_model":"claude-opus-4-1","deprecation_date":"2026-08-05","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"output_format","label":"Output Format","helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","options":[{"label":"Output output","value":"json_schema","subFields":[{"id":"schema","label":"Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object"}}]}]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"claude-opus-4-5-20251101":{"mode":"chat","base_model":"claude-opus-4-5","supports_web_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2026-11-24","max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high"],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"claude-opus-4-5":{"mode":"chat","base_model":"claude-opus-4-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2026-11-24","supports_web_search":true,"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high"],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":false,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider_specific_entry":{"us":1.1,"fast":6},"provider":"anthropic","web_search_billing_unit":"per_request","supports_web_search":true,"deprecation_date":"2027-02-05","inference_geo_us_multiplier":1.1,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":true},"fast/claude-opus-4-6":{"mode":"chat","base_model":"fast/claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true},"us/claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true},"fast/us/claude-opus-4-6":{"mode":"chat","base_model":"fast/us/claude-opus-4-6","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true},"fast/claude-opus-4-6-20260205":{"mode":"chat","base_model":"fast/claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true,"supports_anthropic_compaction":true,"supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_thinking_cache_preservation":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_speed":true},"us/claude-opus-4-6-20260205":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true,"supports_anthropic_compaction":true,"supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_thinking_cache_preservation":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_speed":true},"claude-sonnet-4-20250514":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-06-15","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"cloudflare/@cf/meta/llama-2-7b-chat-fp16":{"mode":"chat","base_model":"","max_input_tokens":3072,"max_output_tokens":3072,"rpm":300,"provider":"cloudflare","max_tokens":3072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":3072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cloudflare/@cf/meta/llama-2-7b-chat-int8":{"mode":"chat","base_model":"","max_input_tokens":2048,"max_output_tokens":2048,"rpm":300,"provider":"cloudflare","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cloudflare/@cf/mistral/mistral-7b-instruct-v0.1":{"mode":"chat","base_model":"","max_input_tokens":8192,"max_output_tokens":8192,"rpm":300,"provider":"cloudflare","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cloudflare/@hf/thebloke/codellama-7b-instruct-awq":{"mode":"chat","base_model":"","max_input_tokens":4096,"max_output_tokens":4096,"rpm":300,"provider":"cloudflare","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-bison":{"mode":"chat","base_model":"code-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-bison-32k@002":{"mode":"completion","base_model":"code-bison-32k","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-bison32k":{"mode":"completion","base_model":"code-bison32k","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-bison@001":{"mode":"completion","base_model":"code-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-bison@002":{"mode":"completion","base_model":"code-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-gecko":{"mode":"completion","base_model":"code-gecko","max_input_tokens":2048,"max_output_tokens":64,"max_tokens":64,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-gecko-latest":{"mode":"completion","base_model":"code-gecko","max_input_tokens":2048,"max_output_tokens":64,"max_tokens":64,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-gecko@001":{"mode":"completion","base_model":"code-gecko","max_input_tokens":2048,"max_output_tokens":64,"max_tokens":64,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"code-gecko@002":{"mode":"completion","base_model":"code-gecko","max_input_tokens":2048,"max_output_tokens":64,"max_tokens":64,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison":{"mode":"chat","base_model":"codechat-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison-32k":{"mode":"chat","base_model":"codechat-bison-32k","max_input_tokens":32000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison-32k@002":{"mode":"chat","base_model":"codechat-bison-32k","max_input_tokens":32000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison@001":{"mode":"chat","base_model":"codechat-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison@002":{"mode":"chat","base_model":"codechat-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codechat-bison@latest":{"mode":"chat","base_model":"codechat-bison","max_input_tokens":6144,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codestral/codestral-2405":{"mode":"chat","base_model":"codestral","deprecation_date":"2025-06-16","deprecation_source":"https://docs.mistral.ai/getting-started/models/","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/capabilities/code_generation/","supports_assistant_prefill":true,"supports_tool_choice":true,"provider":"codestral","is_deprecated":true,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codestral/codestral-latest":{"mode":"chat","base_model":"codestral","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/capabilities/code_generation/","supports_assistant_prefill":true,"supports_tool_choice":true,"provider":"codestral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"codex-mini-latest":{"mode":"responses","base_model":"codex-mini","max_input_tokens":200000,"max_output_tokens":100000,"max_tokens":100000,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cohere.command-light-text-v14":{"mode":"chat","base_model":"command-light-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cohere.command-r-plus-v1:0":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cohere.command-r-v1:0":{"mode":"chat","base_model":"command-r","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cohere.command-text-v14":{"mode":"chat","base_model":"command-text-v14","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","supports_tool_choice":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cohere.embed-english-v3":{"mode":"embedding","base_model":"embed-english-v3","max_input_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_embedding_image_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-west-2","ca-central-1","eu-central-1","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":512},"cohere.embed-multilingual-v3":{"mode":"embedding","base_model":"embed-multilingual-v3","max_input_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_embedding_image_input":true,"provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-west-2","ca-central-1","eu-central-1","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":512},"cohere.embed-v4:0":{"mode":"embedding","base_model":"embed","max_input_tokens":128000,"output_vector_size":1536,"source":"https://aws.amazon.com/bedrock/pricing/","supports_embedding_image_input":true,"provider":"bedrock","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","sa-east-1"],"max_tokens":128000},"cohere/embed-v4.0":{"mode":"embedding","base_model":"embed","max_input_tokens":128000,"output_vector_size":1536,"supports_embedding_image_input":true,"source":"https://cohere.com/pricing","provider":"cohere","max_tokens":128000},"cohere.rerank-v3-5:0":{"mode":"rerank","base_model":"rerank-v3-5","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-rerank-3-5.md","provider":"bedrock","max_document_chunks_per_query":100,"max_query_tokens":32000,"max_tokens":32000,"max_tokens_per_document_chunk":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"command":{"mode":"completion","base_model":"command","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"provider":"cohere","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-a-03-2025":{"mode":"chat","base_model":"command-a","max_input_tokens":256000,"max_output_tokens":8000,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"command-light":{"mode":"chat","base_model":"command-light","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"supports_tool_choice":true,"provider":"cohere","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-nightly":{"mode":"completion","base_model":"command-nightly","max_input_tokens":4096,"max_output_tokens":4096,"provider":"cohere","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-r":{"mode":"chat","base_model":"command-r","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-r-08-2024":{"mode":"chat","base_model":"command-r","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-r-plus":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-r-plus-08-2024":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"command-r7b-12-2024":{"mode":"chat","base_model":"command-r7b","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.cohere.com/v2/docs/command-r7b","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"computer-use-preview":{"mode":"chat","base_model":"computer-use","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":8192,"max_output_tokens":1024,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://platform.openai.com/docs/models/computer-use-preview","provider":"azure","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dall-e-2":{"mode":"image_generation","base_model":"dall-e-2","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/images/variations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek-chat":{"mode":"chat","base_model":"deepseek-chat","max_input_tokens":131072,"max_output_tokens":8192,"max_tokens":8192,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"deepseek","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_native_effort":true},"deepseek-reasoner":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"provider":"deepseek","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_native_effort":true},"dashscope/qwen-coder":{"mode":"chat","base_model":"qwen-coder","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-flash":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-flash-2025-07-28":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-max":{"mode":"chat","base_model":"qwen-max","max_input_tokens":30720,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-2025-01-25":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-2025-04-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-2025-07-14":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-2025-07-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-2025-09-11":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-plus-latest":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-turbo":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-turbo-2024-11-01":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-turbo-2025-04-28":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen-turbo-latest":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-30b-a3b":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"dashscope/qwen3-coder-flash":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"cache_read_input_token_cost":1.2e-7,"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"cache_read_input_token_cost":2e-7,"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"cache_read_input_token_cost":4e-7,"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-coder-flash-2025-07-28":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-coder-plus":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":1e-7,"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"cache_read_input_token_cost":1.8e-7,"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"cache_read_input_token_cost":3e-7,"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"cache_read_input_token_cost":6e-7,"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-coder-plus-2025-07-22":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-max-preview":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwen3-max":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dashscope/qwq-plus":{"mode":"chat","base_model":"qwq-plus","max_input_tokens":98304,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-bge-large-en":{"mode":"embedding","base_model":"bge-large-en","max_input_tokens":512,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"output_vector_size":1024,"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","max_tokens":512},"databricks/databricks-claude-3-7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":4096,"provider":"databricks","supports_vision":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":1024,"provider":"databricks","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-opus-4-1":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":1024,"provider":"databricks","supports_vision":true,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-opus-4-5":{"mode":"chat","base_model":"claude-opus-4-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"provider":"databricks","supports_vision":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-09","successor_model":"databricks/databricks-claude-sonnet-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"prompt_cache_min_tokens":1024,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"provider":"databricks","supports_vision":true,"is_deprecated":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-sonnet-4-1":{"mode":"chat","base_model":"claude-sonnet-4-1","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"provider":"databricks","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":1024,"provider":"databricks","supports_vision":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gemini-2-5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","deprecation_date":"2026-10-02","successor_model":"databricks/databricks-gemini-3-5-flash","max_input_tokens":1048576,"max_output_tokens":65535,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_anthropic_thinking_payload":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","max_tokens":65535,"supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gemini-2-5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_anthropic_thinking_payload":true,"supports_prompt_caching":true,"supports_tool_choice":true,"deprecation_date":"2026-10-02","supports_reasoning":true,"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"is_deprecated":true,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gemma-3-12b":{"mode":"chat","base_model":"gemma-3-12b","max_input_tokens":128000,"max_output_tokens":32000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","max_tokens":32000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"databricks/databricks-gpt-5":{"mode":"chat","base_model":"gpt-5","max_input_tokens":272000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gpt-5-1":{"mode":"chat","base_model":"gpt-5-1","max_input_tokens":272000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","max_input_tokens":272000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","max_input_tokens":272000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"databricks/databricks-gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"databricks/databricks-gte-large-en":{"mode":"embedding","base_model":"gte-large-en","max_input_tokens":8192,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"output_vector_size":1024,"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","max_tokens":8192},"databricks/databricks-llama-2-70b-chat":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"databricks/databricks-llama-4-maverick":{"mode":"chat","base_model":"llama-4-maverick","max_input_tokens":128000,"max_output_tokens":128000,"metadata":{"notes":"Databricks documentation now provides both DBU costs (_dbu_cost_per_token) and dollar costs(_cost_per_token)."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-meta-llama-3-1-405b-instruct":{"mode":"chat","base_model":"llama-3-1-405b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-meta-llama-3-1-8b-instruct":{"mode":"chat","base_model":"llama-3-1-8b-instruct","max_input_tokens":200000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-meta-llama-3-3-70b-instruct":{"mode":"chat","base_model":"llama-3-3-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","max_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_vision":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-meta-llama-3-70b-instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"databricks/databricks-mixtral-8x7b-instruct":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-mpt-30b-instruct":{"mode":"chat","base_model":"mpt-30b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"databricks/databricks-mpt-7b-instruct":{"mode":"chat","base_model":"mpt-7b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_tool_choice":true,"provider":"databricks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dataforseo/search":{"mode":"search","base_model":"search","provider":"dataforseo","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"davinci-002":{"mode":"completion","base_model":"davinci-002","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16384,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","provider":"text-completion-openai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":false,"supports_tool_choice":false,"supports_parallel_function_calling":false,"supports_reasoning":false,"supports_response_schema":false,"supports_prompt_caching":false,"supports_web_search":false,"supports_service_tier":false,"supports_assistant_prefill":false,"supports_system_messages":false},"deepgram/base":{"mode":"audio_transcription","base_model":"base","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-conversationalai":{"mode":"audio_transcription","base_model":"base-conversationalai","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-finance":{"mode":"audio_transcription","base_model":"base-finance","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-general":{"mode":"audio_transcription","base_model":"base-general","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-meeting":{"mode":"audio_transcription","base_model":"base-meeting","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-phonecall":{"mode":"audio_transcription","base_model":"base-phonecall","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-video":{"mode":"audio_transcription","base_model":"base-video","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/base-voicemail":{"mode":"audio_transcription","base_model":"base-voicemail","metadata":{"calculation":"$0.0125/60 seconds = $0.00020833 per second","original_pricing_per_minute":0.0125},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/enhanced":{"mode":"audio_transcription","base_model":"enhanced","metadata":{"calculation":"$0.0145/60 seconds = $0.00024167 per second","original_pricing_per_minute":0.0145},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/enhanced-finance":{"mode":"audio_transcription","base_model":"enhanced-finance","metadata":{"calculation":"$0.0145/60 seconds = $0.00024167 per second","original_pricing_per_minute":0.0145},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/enhanced-general":{"mode":"audio_transcription","base_model":"enhanced-general","metadata":{"calculation":"$0.0145/60 seconds = $0.00024167 per second","original_pricing_per_minute":0.0145},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/enhanced-meeting":{"mode":"audio_transcription","base_model":"enhanced-meeting","metadata":{"calculation":"$0.0145/60 seconds = $0.00024167 per second","original_pricing_per_minute":0.0145},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/enhanced-phonecall":{"mode":"audio_transcription","base_model":"enhanced-phonecall","metadata":{"calculation":"$0.0145/60 seconds = $0.00024167 per second","original_pricing_per_minute":0.0145},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova":{"mode":"audio_transcription","base_model":"nova","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2":{"mode":"audio_transcription","base_model":"nova-2","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-atc":{"mode":"audio_transcription","base_model":"nova-2-atc","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-automotive":{"mode":"audio_transcription","base_model":"nova-2-automotive","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-conversationalai":{"mode":"audio_transcription","base_model":"nova-2-conversationalai","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-drivethru":{"mode":"audio_transcription","base_model":"nova-2-drivethru","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-finance":{"mode":"audio_transcription","base_model":"nova-2-finance","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-general":{"mode":"audio_transcription","base_model":"nova-2-general","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-meeting":{"mode":"audio_transcription","base_model":"nova-2-meeting","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-phonecall":{"mode":"audio_transcription","base_model":"nova-2-phonecall","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-video":{"mode":"audio_transcription","base_model":"nova-2-video","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-2-voicemail":{"mode":"audio_transcription","base_model":"nova-2-voicemail","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-3":{"mode":"audio_transcription","base_model":"nova-3","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-3-general":{"mode":"audio_transcription","base_model":"nova-3-general","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-3-medical":{"mode":"audio_transcription","base_model":"nova-3-medical","metadata":{"calculation":"$0.0052/60 seconds = $0.00008667 per second (multilingual)","original_pricing_per_minute":0.0052},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-general":{"mode":"audio_transcription","base_model":"nova-general","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/nova-phonecall":{"mode":"audio_transcription","base_model":"nova-phonecall","metadata":{"calculation":"$0.0043/60 seconds = $0.00007167 per second","original_pricing_per_minute":0.0043},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper":{"mode":"audio_transcription","base_model":"whisper","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper-base":{"mode":"audio_transcription","base_model":"whisper-base","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper-large":{"mode":"audio_transcription","base_model":"whisper-large","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper-medium":{"mode":"audio_transcription","base_model":"whisper-medium","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper-small":{"mode":"audio_transcription","base_model":"whisper-small","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepgram/whisper-tiny":{"mode":"audio_transcription","base_model":"whisper-tiny","metadata":{"notes":"Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models"},"source":"https://deepgram.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"deepgram","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Gryphe/MythoMax-L2-13b":{"mode":"chat","base_model":"mythomax-l2-13b","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/NousResearch/Hermes-3-Llama-3.1-405B":{"mode":"chat","base_model":"hermes-3-llama-3.1-405b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/NousResearch/Hermes-3-Llama-3.1-70B":{"mode":"chat","base_model":"hermes-3-llama-3.1-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/QwQ-32B":{"mode":"chat","base_model":"qwq-32b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-72b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen2.5-7B-Instruct":{"mode":"chat","base_model":"qwen2.5-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen2.5-VL-32B-Instruct":{"mode":"chat","base_model":"qwen2.5-vl-32b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_tool_choice":true,"supports_vision":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-14B":{"mode":"chat","base_model":"qwen3-14b","max_input_tokens":40960,"max_output_tokens":40960,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-235B-A22B":{"mode":"chat","base_model":"qwen3-235b-a22b","max_input_tokens":40960,"max_output_tokens":40960,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-30B-A3B":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":40960,"max_output_tokens":40960,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-32B":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":40960,"max_output_tokens":40960,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct-turbo","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"supports_prompt_caching":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Qwen/Qwen3-Next-80B-A3B-Thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Sao10K/L3-8B-Lunaris-v1-Turbo":{"mode":"chat","base_model":"l3-8b-lunaris-v1-turbo","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Sao10K/L3.1-70B-Euryale-v2.2":{"mode":"chat","base_model":"l3.1-70b-euryale","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/Sao10K/L3.3-70B-Euryale-v2.3":{"mode":"chat","base_model":"l3.3-70b-euryale","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/allenai/olmOCR-7B-0725-FP8":{"mode":"chat","base_model":"olmocr-7b-0725-fp8","max_input_tokens":16384,"max_output_tokens":16384,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/anthropic/claude-3-7-sonnet-latest":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":200000,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/anthropic/claude-4-opus":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":200000,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/anthropic/claude-4-sonnet":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":200000,"max_output_tokens":200000,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1-0528-Turbo":{"mode":"chat","base_model":"deepseek-r1-0528-turbo","max_input_tokens":32768,"max_output_tokens":32768,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-32b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-R1-Turbo":{"mode":"chat","base_model":"deepseek-r1-turbo","max_input_tokens":40960,"max_output_tokens":40960,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-V3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_function_calling":true,"supports_prompt_caching":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"deepseek-v3.1","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":163840,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V3.1-Terminus":{"mode":"chat","base_model":"deepseek-v3.1-terminus","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/google/gemini-2.0-flash-001":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_tokens":1000000,"max_input_tokens":1000000,"max_output_tokens":1000000,"supports_tool_choice":true,"provider":"deepinfra","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"deepinfra/google/gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","max_input_tokens":1000000,"max_output_tokens":1000000,"supports_tool_choice":true,"supports_function_calling":true,"supports_image_size":false,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/google/gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":1000000,"max_output_tokens":1000000,"supports_tool_choice":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/google/gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/google/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/google/gemma-3-4b-it":{"mode":"chat","base_model":"gemma-3-4b-it","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct":{"mode":"chat","base_model":"llama-3.2-11b-vision-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-3.2-3B-Instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-3.3-70B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.3-70b-instruct-turbo","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":327680,"max_output_tokens":327680,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":327680,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":327680},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-Guard-3-8B":{"mode":"chat","base_model":"llama-guard-3-8b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/meta-llama/Llama-Guard-4-12B":{"mode":"chat","base_model":"llama-guard-4-12b","max_input_tokens":163840,"max_output_tokens":163840,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Meta-Llama-3-8B-Instruct":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Meta-Llama-3.1-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.1-70b-instruct-turbo","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Meta-Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.1-8b-instruct-turbo","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/microsoft/WizardLM-2-8x22B":{"mode":"chat","base_model":"wizardlm-2-8x22b","max_input_tokens":65536,"max_output_tokens":65536,"supports_tool_choice":false,"provider":"deepinfra","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/microsoft/phi-4":{"mode":"chat","base_model":"phi-4","max_input_tokens":16384,"max_output_tokens":16384,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/mistralai/Mistral-Nemo-Instruct-2407":{"mode":"chat","base_model":"mistral-nemo-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/mistralai/Mistral-Small-24B-Instruct-2501":{"mode":"chat","base_model":"mistral-small-24b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/mistralai/Mistral-Small-3.2-24B-Instruct-2506":{"mode":"chat","base_model":"mistral-small-3.2-24b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/moonshotai/Kimi-K2-Instruct":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/moonshotai/Kimi-K2-Instruct-0905":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-nemotron-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/nvidia/Llama-3.3-Nemotron-Super-49B-v1.5":{"mode":"chat","base_model":"llama-3.3-nemotron-super-49b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/nvidia/NVIDIA-Nemotron-Nano-9B-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-9b-v2","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepinfra/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"deepinfra/zai-org/GLM-4.5":{"mode":"chat","base_model":"glm-4.5","max_input_tokens":131072,"max_output_tokens":131072,"supports_tool_choice":true,"supports_function_calling":true,"provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-chat":{"mode":"chat","base_model":"deepseek-chat","deprecation_date":"2026-07-24","deprecation_source":"https://api-docs.deepseek.com/updates","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"deepseek","supports_native_effort":true,"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-coder":{"mode":"chat","base_model":"deepseek-coder","max_input_tokens":128000,"max_output_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"deepseek","supports_native_effort":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":65536,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"deepseek","supports_native_effort":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-reasoner":{"mode":"chat","base_model":"deepseek-r1","deprecation_date":"2026-07-24","deprecation_source":"https://api-docs.deepseek.com/updates","max_input_tokens":131072,"max_output_tokens":65536,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"provider":"deepseek","supports_native_effort":true,"is_deprecated":true,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-v3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":65536,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"deepseek","supports_native_effort":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek/deepseek-v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"deepseek","supports_native_effort":true,"max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek.v3-v1:0":{"mode":"chat","base_model":"deepseek-v3","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html","max_input_tokens":163840,"max_output_tokens":81920,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_native_structured_output":true,"provider":"bedrock","max_tokens":81920,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":81920},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"dolphin":{"mode":"completion","base_model":"dolphin","max_input_tokens":16384,"max_output_tokens":16384,"provider":"nlp_cloud","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"deepseek-v3-2-251201":{"mode":"chat","base_model":"deepseek-v3-2","max_input_tokens":98304,"max_output_tokens":32768,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"volcengine","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"glm-4-7-251222":{"mode":"chat","base_model":"glm-4-7","max_input_tokens":204800,"max_output_tokens":131072,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"volcengine","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"kimi-k2-thinking-251104":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":229376,"max_output_tokens":32768,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"volcengine","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"doubao-embedding":{"mode":"embedding","base_model":"doubao-embedding","max_input_tokens":4096,"metadata":{"notes":"Volcengine Doubao embedding model - standard version with 2560 dimensions"},"output_vector_size":2560,"provider":"volcengine","max_tokens":4096},"doubao-embedding-large":{"mode":"embedding","base_model":"doubao-embedding-large","max_input_tokens":4096,"metadata":{"notes":"Volcengine Doubao embedding model - large version with 2048 dimensions"},"output_vector_size":2048,"provider":"volcengine","max_tokens":4096},"doubao-embedding-large-text-240915":{"mode":"embedding","base_model":"doubao-embedding-large-text","max_input_tokens":4096,"metadata":{"notes":"Volcengine Doubao embedding model - text-240915 version with 4096 dimensions"},"output_vector_size":4096,"provider":"volcengine","max_tokens":4096},"doubao-embedding-large-text-250515":{"mode":"embedding","base_model":"doubao-embedding-large-text","max_input_tokens":4096,"metadata":{"notes":"Volcengine Doubao embedding model - text-250515 version with 2048 dimensions"},"output_vector_size":2048,"provider":"volcengine","max_tokens":4096},"doubao-embedding-text-240715":{"mode":"embedding","base_model":"doubao-embedding-text","max_input_tokens":4096,"metadata":{"notes":"Volcengine Doubao embedding model - text-240715 version with 2560 dimensions"},"output_vector_size":2560,"provider":"volcengine","max_tokens":4096},"exa_ai/search":{"mode":"search","base_model":"search","tiered_pricing":[{"input_cost_per_query":0.005,"max_results_range":[0,25]},{"input_cost_per_query":0.025,"max_results_range":[26,100]}],"provider":"exa_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"firecrawl/search":{"mode":"search","base_model":"search","tiered_pricing":[{"input_cost_per_query":0.00166,"max_results_range":[1,10]},{"input_cost_per_query":0.00332,"max_results_range":[11,20]},{"input_cost_per_query":0.00498,"max_results_range":[21,30]},{"input_cost_per_query":0.00664,"max_results_range":[31,40]},{"input_cost_per_query":0.0083,"max_results_range":[41,50]},{"input_cost_per_query":0.00996,"max_results_range":[51,60]},{"input_cost_per_query":0.01162,"max_results_range":[61,70]},{"input_cost_per_query":0.01328,"max_results_range":[71,80]},{"input_cost_per_query":0.01494,"max_results_range":[81,90]},{"input_cost_per_query":0.0166,"max_results_range":[91,100]}],"metadata":{"notes":"Firecrawl search pricing: $83 for 100,000 credits, 2 credits per 10 results. Cost = ceiling(limit/10) * 2 * $0.00083"},"provider":"firecrawl","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/search":{"mode":"search","base_model":"search","provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"searxng/search":{"mode":"search","base_model":"search","metadata":{"notes":"SearXNG is an open-source metasearch engine. Free to use when self-hosted or using public instances."},"provider":"searxng","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"elevenlabs/scribe_v1":{"mode":"audio_transcription","base_model":"scribe-v1","metadata":{"calculation":"$0.22/hour = $0.00366/minute = $0.0000611 per second (enterprise pricing)","notes":"ElevenLabs Scribe v1 - state-of-the-art speech recognition model with 99 language support","original_pricing_per_hour":0.22},"source":"https://elevenlabs.io/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"elevenlabs","model_parameters":[{"id":"language","label":"Language","helpText":"The language of the audio to transcribe. If not specified, the model will auto-detect the language.","type":"select","default":"auto","options":[{"label":"Auto-detect","value":"auto"},{"label":"English","value":"en"},{"label":"Spanish","value":"es"},{"label":"French","value":"fr"},{"label":"German","value":"de"},{"label":"Italian","value":"it"},{"label":"Portuguese","value":"pt"},{"label":"Polish","value":"pl"},{"label":"Dutch","value":"nl"},{"label":"Japanese","value":"ja"},{"label":"Chinese","value":"zh"},{"label":"Korean","value":"ko"},{"label":"Russian","value":"ru"},{"label":"Arabic","value":"ar"},{"label":"Hindi","value":"hi"}]},{"id":"timestamp_granularities","label":"Timestamp Granularities","helpText":"The level of detail for timestamps in the transcription.","type":"select","default":"segment","options":[{"label":"Segment","value":"segment"},{"label":"Word","value":"word"}]},{"id":"diarization","label":"Speaker Diarization","helpText":"Enable speaker diarization to identify and separate different speakers in the audio.","type":"boolean","default":false},{"id":"num_speakers","label":"Number of Speakers","helpText":"Expected number of speakers in the audio (used when diarization is enabled).","type":"number","default":2,"range":{"min":1,"max":10,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the transcription output.","type":"select","default":"json","options":[{"label":"JSON","value":"json"},{"label":"Text","value":"text"},{"label":"SRT","value":"srt"},{"label":"VTT","value":"vtt"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"punctuation","label":"Punctuation","helpText":"Enable automatic punctuation in the transcription.","type":"boolean","default":true},{"id":"remove_disfluencies","label":"Remove Disfluencies","helpText":"Remove filler words like 'um', 'uh', stutters, and false starts from the transcription.","type":"boolean","default":false},{"id":"profanity_filter","label":"Profanity Filter","helpText":"Enable filtering of profane words in the transcription.","type":"boolean","default":false},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"elevenlabs/scribe_v1_experimental":{"mode":"audio_transcription","base_model":"scribe-v1","metadata":{"calculation":"$0.22/hour = $0.00366/minute = $0.0000611 per second (enterprise pricing)","notes":"ElevenLabs Scribe v1 experimental - enhanced version of the main Scribe model","original_pricing_per_hour":0.22},"source":"https://elevenlabs.io/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"elevenlabs","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"elevenlabs/eleven_v3":{"mode":"audio_speech","base_model":"eleven-v3","metadata":{"calculation":"$0.18/1000 characters (Scale plan pricing, 1 credit per character)","notes":"ElevenLabs Eleven v3 - most expressive TTS model with 70+ languages and audio tags support"},"source":"https://elevenlabs.io/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"elevenlabs/eleven_multilingual_v2":{"mode":"audio_speech","base_model":"eleven-multilingual-v2","metadata":{"calculation":"$0.18/1000 characters (Scale plan pricing, 1 credit per character)","notes":"ElevenLabs Eleven Multilingual v2 - default TTS model with 29 languages support"},"source":"https://elevenlabs.io/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","model_parameters":[{"id":"voice_id","label":"Voice","helpText":"The voice to use for speech synthesis. You can use pre-made voices or custom cloned voices.","type":"select","required":true},{"id":"stability","label":"Stability","helpText":"Controls the consistency and predictability of the voice. Lower values allow for more variable and expressive speech, while higher values make the voice more stable and consistent.","type":"number","default":0.5,"range":{"min":0,"max":1,"step":0.01}},{"id":"similarity_boost","label":"Similarity Boost","helpText":"Enhances the similarity to the original voice. Higher values make the generated speech closer to the original voice characteristics.","type":"number","default":0.75,"range":{"min":0,"max":1,"step":0.01}},{"id":"style","label":"Style Exaggeration","helpText":"Controls how much the AI should exaggerate the style of the voice. Higher values result in more expressive and dramatic speech.","type":"number","default":0,"range":{"min":0,"max":1,"step":0.01}},{"id":"use_speaker_boost","label":"Speaker Boost","helpText":"Boost the similarity to the speaker and reduces differences between speakers. Useful when using different voices in the same audio.","type":"boolean","default":true},{"id":"model_id","label":"Model","helpText":"The TTS model to use for generation.","type":"select","default":"eleven_multilingual_v2","options":[{"label":"Eleven Multilingual v2","value":"eleven_multilingual_v2"},{"label":"Eleven Turbo v2","value":"eleven_turbo_v2"},{"label":"Eleven Turbo v2.5","value":"eleven_turbo_v2_5"},{"label":"Eleven Monolingual v1","value":"eleven_monolingual_v1"}]},{"id":"language_code","label":"Language","helpText":"The language code for the text to be synthesized. Required for multilingual models to optimize pronunciation.","type":"select","default":"en","options":[{"label":"English","value":"en"},{"label":"Spanish","value":"es"},{"label":"French","value":"fr"},{"label":"German","value":"de"},{"label":"Italian","value":"it"},{"label":"Portuguese","value":"pt"},{"label":"Polish","value":"pl"},{"label":"Dutch","value":"nl"},{"label":"Japanese","value":"ja"},{"label":"Chinese","value":"zh"},{"label":"Korean","value":"ko"},{"label":"Russian","value":"ru"},{"label":"Arabic","value":"ar"},{"label":"Hindi","value":"hi"},{"label":"Turkish","value":"tr"},{"label":"Swedish","value":"sv"},{"label":"Indonesian","value":"id"},{"label":"Filipino","value":"fil"},{"label":"Ukrainian","value":"uk"},{"label":"Greek","value":"el"},{"label":"Czech","value":"cs"},{"label":"Finnish","value":"fi"},{"label":"Romanian","value":"ro"},{"label":"Danish","value":"da"},{"label":"Bulgarian","value":"bg"},{"label":"Malay","value":"ms"},{"label":"Slovak","value":"sk"},{"label":"Croatian","value":"hr"},{"label":"Tamil","value":"ta"},{"label":"Vietnamese","value":"vi"}]},{"id":"output_format","label":"Output Format","helpText":"The audio format for the generated speech.","type":"select","default":"mp3_44100_128","options":[{"label":"MP3 (44.1kHz, 128kbps)","value":"mp3_44100_128"},{"label":"MP3 (44.1kHz, 192kbps)","value":"mp3_44100_192"},{"label":"MP3 (22.05kHz, 32kbps)","value":"mp3_22050_32"},{"label":"PCM (16kHz, 16-bit)","value":"pcm_16000"},{"label":"PCM (22.05kHz, 16-bit)","value":"pcm_22050"},{"label":"PCM (24kHz, 16-bit)","value":"pcm_24000"},{"label":"PCM (44.1kHz, 16-bit)","value":"pcm_44100"},{"label":"��-law (8kHz, 8-bit)","value":"ulaw_8000"}]},{"id":"optimize_streaming_latency","label":"Optimize Streaming Latency","helpText":"Optimize the model for lower latency at the cost of quality. Higher values reduce latency but may decrease audio quality.","type":"number","default":0,"range":{"min":0,"max":4,"step":1}},{"id":"seed","label":"Seed","helpText":"Random seed for reproducible generation. If set, the same text with the same settings will produce identical audio.","type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"embed-english-light-v2.0":{"mode":"embedding","base_model":"embed-english-light","max_input_tokens":1024,"max_tokens":1024,"provider":"cohere","deprecation_date":"2026-04-04","is_deprecated":true},"embed-english-light-v3.0":{"mode":"embedding","base_model":"embed-english-light","max_input_tokens":1024,"provider":"cohere","max_tokens":1024},"embed-english-v2.0":{"mode":"embedding","base_model":"embed-english","max_input_tokens":4096,"max_tokens":4096,"provider":"cohere","deprecation_date":"2026-04-04","is_deprecated":true},"embed-english-v3.0":{"mode":"embedding","base_model":"embed-english","max_input_tokens":1024,"metadata":{"notes":"'supports_image_input' is a deprecated field. Use 'supports_embedding_image_input' instead."},"supports_embedding_image_input":true,"supports_image_input":true,"provider":"cohere","max_tokens":1024},"embed-multilingual-v2.0":{"mode":"embedding","base_model":"embed-multilingual","max_input_tokens":768,"max_tokens":768,"provider":"cohere","deprecation_date":"2026-04-04","is_deprecated":true},"embed-multilingual-v3.0":{"mode":"embedding","base_model":"embed-multilingual","max_input_tokens":1024,"supports_embedding_image_input":true,"provider":"cohere","max_tokens":1024},"embed-multilingual-light-v3.0":{"mode":"embedding","base_model":"embed-multilingual-light","max_input_tokens":1024,"supports_embedding_image_input":true,"provider":"cohere","max_tokens":1024},"eu.amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.amazon.nova-micro-v1:0":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":10000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-5-haiku-20241022-v1:0":{"mode":"chat","base_model":"claude-3-5-haiku","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"prompt_cache_min_tokens":2048,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2026-10-15","successor_model":"eu.anthropic.claude-haiku-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":4096,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_cache_point":false,"is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-5-sonnet-20241022-v2:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-opus-20240229-v1:0":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_cache_point":false,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-3-sonnet-20240229-v1:0":{"mode":"chat","base_model":"claude-3-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-opus-4-1-20250805-v1:0":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"deprecation_date":"2027-01-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","token_count_model":{"model":"anthropic.claude-opus-4-1-20250805-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":32000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-opus-4-20250514-v1:0":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"bedrock","max_tokens":32000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-14","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"eu.anthropic.claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"bedrock_converse_supports_strict_tools":false,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-20250514-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1"],"max_tokens":64000,"is_deprecated":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"eu.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"eu.meta.llama3-2-1b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-1b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.meta.llama3-2-3b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-3b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.mistral.pixtral-large-2502-v1:0":{"mode":"chat","base_model":"pixtral-large","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-pixtral-large.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/bria/text-to-image/3.2":{"mode":"image_generation","base_model":"bria/text-to-image/3.2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/flux-pro/v1.1":{"mode":"image_generation","base_model":"flux-pro/v1.1","metadata":{"notes":"fal bills fal-ai/flux-pro/v1.1 at $0.04 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"source":"https://fal.ai/models/fal-ai/flux-pro/v1.1","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/flux-pro/v1.1-ultra":{"mode":"image_generation","base_model":"flux-pro/v1.1-ultra","source":"https://fal.ai/models/fal-ai/flux-pro/v1.1-ultra","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/flux/schnell":{"mode":"image_generation","base_model":"flux/schnell","metadata":{"notes":"fal bills fal-ai/flux/schnell at $0.003 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"source":"https://fal.ai/models/fal-ai/flux/schnell","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/bytedance/seedream/v3/text-to-image":{"mode":"image_generation","base_model":"seedream/v3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/bytedance/dreamina/v3.1/text-to-image":{"mode":"image_generation","base_model":"dreamina/v3.1/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/ideogram/v3":{"mode":"image_generation","base_model":"ideogram/v3","source":"https://fal.ai/models/fal-ai/ideogram/v3","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/imagen4/preview":{"mode":"image_generation","base_model":"imagen4/preview","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/imagen4/preview/fast":{"mode":"image_generation","base_model":"imagen4/preview/fast","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/imagen4/preview/ultra":{"mode":"image_generation","base_model":"imagen4/preview/ultra","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/recraft/v3/text-to-image":{"mode":"image_generation","base_model":"v3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fal_ai/fal-ai/stable-diffusion-v35-medium":{"mode":"image_generation","base_model":"stable-diffusion-v35-medium","metadata":{"notes":"fal bills fal-ai/stable-diffusion-v35-medium at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"source":"https://fal.ai/models/fal-ai/stable-diffusion-v35-medium","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"featherless_ai/featherless-ai/Qwerky-72B":{"mode":"chat","base_model":"qwerky-72b","max_input_tokens":32768,"max_output_tokens":4096,"provider":"featherless_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"featherless_ai/featherless-ai/Qwerky-QwQ-32B":{"mode":"chat","base_model":"qwerky-qwq-32b","max_input_tokens":32768,"max_output_tokens":4096,"provider":"featherless_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-4.1b-to-16b":{"mode":"chat","base_model":"fireworks-ai-4.1b-to-16b","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-56b-to-176b":{"mode":"chat","base_model":"fireworks-ai-56b-to-176b","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-above-16b":{"mode":"chat","base_model":"fireworks-ai-above-16b","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-default":{"mode":"chat","base_model":"fireworks-ai-default","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-embedding-150m-to-350m":{"mode":"chat","base_model":"fireworks-ai-embedding-150m-to-350m","provider":"fireworks_ai-embedding-models","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-embedding-up-to-150m":{"mode":"chat","base_model":"fireworks-ai-embedding-up-to-150m","provider":"fireworks_ai-embedding-models","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-moe-up-to-56b":{"mode":"chat","base_model":"fireworks-ai-moe-up-to-56b","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks-ai-up-to-4b":{"mode":"chat","base_model":"fireworks-ai-up-to-4b","source":"https://docs.fireworks.ai/serverless/pricing","provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/WhereIsAI/UAE-Large-V1":{"mode":"embedding","base_model":"uae-large-v1","max_input_tokens":512,"source":"https://fireworks.ai/pricing","provider":"fireworks_ai-embedding-models","max_tokens":512},"fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-instruct":{"mode":"chat","base_model":"deepseek-coder-v2-instruct","max_input_tokens":65536,"max_output_tokens":65536,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/deepseek-r1","max_input_tokens":160000,"max_output_tokens":160000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":false,"provider":"litellm","max_tokens":160000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":160000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true},"fireworks_ai/accounts/fireworks/models/deepseek-r1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":160000,"max_output_tokens":160000,"source":"https://fireworks.ai/pricing","supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":160000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":160000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-basic":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/deepseek-r1-basic","max_input_tokens":160000,"max_output_tokens":160000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":false,"provider":"litellm","max_tokens":160000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":160000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true},"fireworks_ai/accounts/fireworks/models/deepseek-v3":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/deepseek-v3","max_input_tokens":128000,"max_output_tokens":128000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":true,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true},"fireworks_ai/accounts/fireworks/models/deepseek-v3-0324":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/deepseek-v3-0324","max_input_tokens":160000,"max_output_tokens":160000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":true,"provider":"litellm","max_tokens":160000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":160000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true},"fireworks_ai/accounts/fireworks/models/deepseek-v3p1":{"mode":"chat","base_model":"deepseek-v3.1","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://fireworks.ai/pricing","supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v3p1-terminus":{"mode":"chat","base_model":"deepseek-v3.1-terminus","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://fireworks.ai/pricing","supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v3p2":{"mode":"chat","base_model":"deepseek-v3.2","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://fireworks.ai/models/fireworks/deepseek-v3p2","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/firefunction-v2":{"mode":"chat","base_model":"firefunction-v2","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/glm-4p5":{"mode":"chat","base_model":"glm-4.5","max_input_tokens":128000,"max_output_tokens":96000,"source":"https://fireworks.ai/models/fireworks/glm-4p5","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":96000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/glm-4p5-air":{"mode":"chat","base_model":"glm-4.5-air","max_input_tokens":128000,"max_output_tokens":96000,"source":"https://artificialanalysis.ai/models/glm-4-5-air","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":96000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/glm-4p6":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":202800,"max_output_tokens":202800,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":202800,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":202800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":131072,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2-instruct":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":131072,"max_output_tokens":16384,"source":"https://fireworks.ai/models/fireworks/kimi-k2-instruct","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2-instruct-0905":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://app.fireworks.ai/models/fireworks/kimi-k2-instruct-0905","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2p5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p1-405b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-v3p1-405b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p1-8b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-v3p1-8b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-11b-vision-instruct":{"mode":"chat","base_model":"llama-3.2-11b-vision-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-1b-instruct":{"mode":"chat","base_model":"llama-3.2-1b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-3b-instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-90b-vision-instruct":{"mode":"chat","base_model":"llama-3.2-90b-vision-instruct","max_input_tokens":16384,"max_output_tokens":16384,"source":"https://fireworks.ai/pricing","supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama4-maverick-instruct-basic":{"mode":"image_generation","base_model":"fireworks/accounts/fireworks/models/llama4-maverick-instruct-basic","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":false,"provider":"litellm","max_tokens":1000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/llama4-scout-instruct-basic":{"mode":"image_generation","base_model":"fireworks/accounts/fireworks/models/llama4-scout-instruct-basic","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_tool_choice":false,"provider":"litellm","max_tokens":1000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/mixtral-8x22b-instruct-hf":{"mode":"chat","base_model":"mixtral-8x22b-instruct","max_input_tokens":65536,"max_output_tokens":65536,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2-72b-instruct":{"mode":"chat","base_model":"qwen2-72b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/yi-large":{"mode":"chat","base_model":"yi-large","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://fireworks.ai/pricing","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/nomic-ai/nomic-embed-text-v1":{"mode":"embedding","base_model":"nomic-embed-text-v1","max_input_tokens":8192,"source":"https://fireworks.ai/pricing","provider":"fireworks_ai-embedding-models","max_tokens":8192},"fireworks_ai/nomic-ai/nomic-embed-text-v1.5":{"mode":"embedding","base_model":"nomic-embed-text","max_input_tokens":8192,"source":"https://fireworks.ai/pricing","provider":"fireworks_ai-embedding-models","max_tokens":8192},"fireworks_ai/thenlper/gte-base":{"mode":"embedding","base_model":"gte-base","max_input_tokens":512,"source":"https://fireworks.ai/pricing","provider":"fireworks_ai-embedding-models","max_tokens":512},"fireworks_ai/thenlper/gte-large":{"mode":"embedding","base_model":"gte-large","max_input_tokens":512,"source":"https://fireworks.ai/pricing","provider":"fireworks_ai-embedding-models","max_tokens":512},"friendliai/meta-llama-3.1-70b-instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"friendliai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"friendliai/meta-llama-3.1-8b-instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"friendliai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:babbage-002":{"mode":"completion","base_model":"babbage-002","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16384,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","provider":"text-completion-openai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:davinci-002":{"mode":"completion","base_model":"davinci-002","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16384,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","provider":"text-completion-openai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-3.5-turbo-0125":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-3.5-turbo-0613":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"ft:gpt-3.5-turbo","max_input_tokens":4096,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-3.5-turbo-1106":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-4-0613":{"mode":"chat","base_model":"gpt-4","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":8192,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"OpenAI needs to add pricing for this ft model, will be updated when added by OpenAI. Defaulting to base model pricing","supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ft:gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:gpt-4.1-2025-04-14":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:gpt-4.1-mini-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:gpt-4.1-nano-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ft:o4-mini-2025-04-16":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini-1.0-pro":{"mode":"chat","base_model":"gemini-1.0-pro","max_input_tokens":32760,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#google_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-1.0-pro-001":{"mode":"chat","base_model":"gemini-1.0-pro","deprecation_date":"2025-04-09","max_input_tokens":32760,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.0-pro-002":{"mode":"chat","base_model":"gemini-1.0-pro","deprecation_date":"2025-04-09","max_input_tokens":32760,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.0-pro-vision":{"mode":"chat","base_model":"gemini-1.0-pro-vision","max_images_per_prompt":16,"max_input_tokens":16384,"max_output_tokens":2048,"max_tokens":2048,"max_video_length":2,"max_videos_per_prompt":1,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-1.0-pro-vision-001":{"mode":"chat","base_model":"gemini-1.0-pro-vision","deprecation_date":"2025-04-09","max_images_per_prompt":16,"max_input_tokens":16384,"max_output_tokens":2048,"max_tokens":2048,"max_video_length":2,"max_videos_per_prompt":1,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.0-ultra":{"mode":"chat","base_model":"gemini-1.0-ultra","max_input_tokens":8192,"max_output_tokens":2048,"max_tokens":2048,"source":"As of Jun, 2024. There is no available doc on vertex ai pricing gemini-1.0-ultra-001. Using gemini-1.0-pro pricing. Got max_tokens info here: https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-1.0-ultra-001":{"mode":"chat","base_model":"gemini-1.0-ultra","max_input_tokens":8192,"max_output_tokens":2048,"max_tokens":2048,"source":"As of Jun, 2024. There is no available doc on vertex ai pricing gemini-1.0-ultra-001. Using gemini-1.0-pro pricing. Got max_tokens info here: https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-1.5-flash":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1000000,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-flash-001":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-05-24","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1000000,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-flash-002":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-24","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-1.5-flash","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-flash-exp-0827":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1000000,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-flash-preview-0514":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1000000,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro-001":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-05-24","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro-002":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-24","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-1.5-pro","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro-preview-0215":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro-preview-0409":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-1.5-pro-preview-0514":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/pricing#2_0flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-001":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-exp":{"mode":"chat","base_model":"gemini-2.0-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.0-flash-lite":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":50,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-lite-001":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":50,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-live-preview-04-09":{"mode":"chat","base_model":"gemini-2.0-flash-live","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://cloud.google.com/vertex-ai/docs/generative-ai/model-reference/gemini#gemini-2-0-flash-live-preview-04-09","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.0-flash-preview-image-generation":{"mode":"chat","base_model":"gemini-2.0-flash-image-generation","deprecation_date":"2025-11-14","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/pricing#2_0flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-thinking-exp":{"mode":"chat","base_model":"gemini-2.0-flash-thinking","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-flash-thinking-exp-01-21":{"mode":"chat","base_model":"gemini-2.0-flash-thinking","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65536,"max_pdf_size_mb":30,"max_tokens":65536,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.0-pro-exp-02-05":{"mode":"chat","base_model":"gemini-2.0-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":2097152,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","deprecation_date":"2026-10-20","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":1048576,"max_output_tokens":65535,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","successor_model":"gemini-3.8-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_image_size":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","is_deprecated":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":1,"max":20,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":false,"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":24576},"supports_response_schema_with_tools":false},"gemini-2.5-flash-image":{"mode":"image_generation","base_model":"gemini-2.5-flash-image","deprecation_date":"2027-03-15","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"gemini-3.1-flash-lite-image","max_input_tokens":32768,"max_output_tokens":32768,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-flash-image","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":false,"tpm":8000000,"supports_image_size":false,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_tokens":32768,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-image-preview":{"mode":"image_generation","base_model":"gemini-2.5-flash-image","deprecation_date":"2026-01-15","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-3-pro-image-preview":{"mode":"image_generation","base_model":"gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai-language-models","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"service_tiers":["flex"]},"deep-research-pro-preview-12-2025":{"mode":"image_generation","base_model":"deep-research-pro","max_input_tokens":65536,"max_output_tokens":32768,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-lite":{"mode":"chat","base_model":"gemini-2.5-flash-lite","deprecation_date":"2026-10-20","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":1048576,"max_output_tokens":65535,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","successor_model":"gemini-3.8-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_image_size":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","is_deprecated":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":false,"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":512,"max":24576},"supports_response_schema_with_tools":false},"gemini-2.5-flash-lite-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"supports_image_size":false,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"supports_image_size":false,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-live-2.5-flash-preview-native-audio-09-2025":{"mode":"chat","base_model":"gemini-live-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"gemini_native_audio":true,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-live-2.5-flash-preview-native-audio-09-2025":{"mode":"chat","base_model":"gemini-live-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"gemini_native_audio":true,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-lite-preview-06-17":{"mode":"chat","base_model":"gemini-2.5-flash-lite","deprecation_date":"2025-11-18","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.5-flash-preview-04-17":{"mode":"chat","base_model":"gemini-2.5-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-preview-05-20":{"mode":"chat","base_model":"gemini-2.5-flash","deprecation_date":"2025-11-18","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","deprecation_date":"2026-10-20","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":1048576,"max_output_tokens":65535,"prompt_cache_min_tokens":2048,"search_free_requests":{"count":10000,"period":"day","shared_group":"gemini-2.5-pro"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","successor_model":"gemini-3.8-flash","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","is_deprecated":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":false,"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":128,"max":32768},"supports_response_schema_with_tools":false},"gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"provider":"vertex_ai-language-models","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"thinking_level","label":"Thinking Level","helpText":"Set the thinking level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_multimodal_tool_output":true,"supports_response_schema_with_tools":true},"vertex_ai/gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_multimodal_tool_output":true,"supports_response_schema_with_tools":true},"vertex_ai/gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65535,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_url_context":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"service_tiers":["priority","flex"],"supports_multimodal_tool_output":true,"supports_response_schema_with_tools":true},"gemini-2.5-pro-exp-03-25":{"mode":"chat","base_model":"gemini-2.5-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-pro-preview-03-25":{"mode":"chat","base_model":"gemini-2.5-pro","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.5-pro-preview-05-06":{"mode":"chat","base_model":"gemini-2.5-pro","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supported_regions":["global"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.5-pro-preview-06-05":{"mode":"chat","base_model":"gemini-2.5-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-pro-preview-tts":{"mode":"chat","base_model":"gemini-2.5-pro-tts","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-robotics-er-1.5-preview":{"mode":"chat","base_model":"gemini-robotics-er-1.5","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-robotics-er-1-5-preview","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":65535,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-robotics-er-1.5-preview":{"mode":"chat","base_model":"gemini-robotics-er-1.5","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-robotics-er-1-5-preview","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"rpm":10,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-computer-use-preview-10-2025":{"mode":"chat","base_model":"gemini-2.5-computer-use","max_input_tokens":128000,"max_output_tokens":64000,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/computer-use","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_images_per_prompt":3000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-embedding-001":{"mode":"embedding","base_model":"gemini-embedding-001","deprecation_date":"2028-05-20","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"gemini-embedding-2","max_input_tokens":2048,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","provider":"vertex_ai","max_tokens":2048},"gemini-flash-experimental":{"mode":"chat","base_model":"gemini-flash","max_input_tokens":1000000,"max_output_tokens":8192,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/gemini-experimental","uses_embed_content":true,"provider":"vertex_ai","max_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-pro":{"mode":"chat","base_model":"gemini-pro","max_input_tokens":32760,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-pro-experimental":{"mode":"chat","base_model":"gemini-pro","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/gemini-experimental","supports_function_calling":false,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-pro-vision":{"mode":"chat","base_model":"gemini-pro-vision","max_images_per_prompt":16,"max_input_tokens":16384,"max_output_tokens":2048,"max_tokens":2048,"max_video_length":2,"max_videos_per_prompt":1,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-embedding-001":{"mode":"embedding","base_model":"gemini-embedding-001","deprecation_date":"2028-05-14","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-embedding-2","max_input_tokens":2048,"output_vector_size":3072,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/embeddings#model-versions","tpm":10000000,"provider":"gemini","max_tokens":2048},"gemini/gemini-1.5-flash":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-001":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-05-24","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-002":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-24","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-8b":{"mode":"chat","base_model":"gemini-1.5-flash-8b","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":4000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-8b-exp-0827":{"mode":"chat","base_model":"gemini-1.5-flash-8b","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1000000,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":4000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-8b-exp-0924":{"mode":"chat","base_model":"gemini-1.5-flash-8b","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":4000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-exp-0827":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-flash-latest":{"mode":"chat","base_model":"gemini-1.5-flash","deprecation_date":"2025-09-29","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro-001":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-05-24","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro-002":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-24","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro-exp-0801":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro-exp-0827":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":2097152,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-1.5-pro-latest":{"mode":"chat","base_model":"gemini-1.5-pro","deprecation_date":"2025-09-29","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/pricing#2_0flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-001":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/pricing#2_0flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-exp":{"mode":"chat","base_model":"gemini-2.0-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.0-flash-lite":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":50,"max_video_length":1,"max_videos_per_prompt":10,"rpm":4000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.0-flash-lite","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-lite-preview-02-05":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":60000,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash-lite","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-live-001":{"mode":"chat","base_model":"gemini-2.0-flash-live","deprecation_date":"2025-12-09","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2-0-flash-live-001","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-preview-image-generation":{"mode":"chat","base_model":"gemini-2.0-flash-image-generation","deprecation_date":"2025-11-14","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/pricing#2_0flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-thinking-exp":{"mode":"chat","base_model":"gemini-2.0-flash-thinking","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65536,"max_pdf_size_mb":30,"max_tokens":65536,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-flash-thinking-exp-01-21":{"mode":"chat","base_model":"gemini-2.0-flash-thinking","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65536,"max_pdf_size_mb":30,"max_tokens":65536,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.0-pro-exp-02-05":{"mode":"chat","base_model":"gemini-2.0-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":2097152,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":1000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","deprecation_date":"2026-12-01","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"supports_audio_input":true,"supports_image_size":false,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":1,"max":20,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-image":{"mode":"image_generation","base_model":"gemini-2.5-flash-image","deprecation_date":"2027-03-15","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.1-flash-lite-image","supports_reasoning":false,"max_input_tokens":32768,"max_output_tokens":32768,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-flash-image","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"supports_audio_input":false,"supports_image_size":false,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_tokens":32768,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-image-preview":{"mode":"image_generation","base_model":"gemini-2.5-flash-image","deprecation_date":"2026-01-15","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-3-pro-image-preview":{"mode":"image_generation","base_model":"gemini-3-pro-image","deprecation_date":"2026-06-25","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","supports_reasoning":false,"provider":"gemini","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/deep-research-pro-preview-12-2025":{"mode":"image_generation","base_model":"deep-research-pro","deprecation_date":"2026-10-23","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/deep-research-preview-04-2026","max_input_tokens":65536,"max_output_tokens":32768,"rpm":1000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model":"deep-research-pro-preview-12-2025","max_tokens":32768,"web_search_billing_unit":"per_query","supports_url_context":true,"supports_code_execution":true,"supports_native_structured_output":false,"supports_native_streaming":true,"supports_reasoning":true,"supports_image_input":true,"supports_audio_input":true,"supports_video_input":true,"supports_multimodal":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-lite":{"mode":"chat","base_model":"gemini-2.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":15,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"search_free_requests":{"count":1500,"period":"day","shared_group":"gemini-2-flash"},"supports_audio_input":true,"supports_image_size":false,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-lite-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash-lite","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":15,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_image_size":false},"gemini/gemini-2.5-flash-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":15,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_image_size":false},"gemini/gemini-flash-latest":{"mode":"chat","base_model":"gemini-flash","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":15,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"prompt_cache_min_tokens":4096,"supports_audio_input":true,"supports_native_streaming":true,"supports_video_input":true,"web_search_billing_unit":"per_query","provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-flash-lite-latest":{"mode":"chat","base_model":"gemini-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":15,"source":"https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"supports_audio_input":true,"supports_native_streaming":true,"supports_video_input":true,"web_search_billing_unit":"per_query","provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-lite-preview-06-17":{"mode":"chat","base_model":"gemini-2.5-flash-lite","deprecation_date":"2025-11-18","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":15,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.5-flash-preview-04-17":{"mode":"chat","base_model":"gemini-2.5-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-preview-05-20":{"mode":"chat","base_model":"gemini-2.5-flash","deprecation_date":"2025-11-18","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.5-flash-preview-tts":{"mode":"audio_speech","base_model":"gemini-2.5-flash-tts","deprecation_date":"2026-12-18","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-flash-tts","max_input_tokens":8192,"max_output_tokens":16384,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/audio/speech"],"tpm":4000000,"rpm":10,"supports_audio_input":false,"supports_function_calling":false,"supports_response_schema":false,"supports_web_search":false,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":2000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"prompt_cache_min_tokens":2048,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"search_free_requests":{"count":10000,"period":"day","shared_group":"gemini-2.5-pro"},"supports_url_context":true,"deprecation_date":"2026-12-01","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-flash","provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gemini/gemini-2.5-computer-use-preview-10-2025":{"mode":"chat","base_model":"gemini-2.5-computer-use","deprecation_date":"2026-07-28","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-flash","max_input_tokens":128000,"max_output_tokens":64000,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/computer-use","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":800000,"provider":"gemini","is_deprecated":true,"max_images_per_prompt":3000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"thinking_level","label":"Thinking Level","helpText":"Set the thinking level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_multimodal_tool_output":true,"supports_response_schema_with_tools":true},"gemini/gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":2000,"source":"https://ai.google.dev/pricing/gemini-3","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","successor_model":"gemini/gemini-3.6-flash","supports_audio_input":true,"supports_computer_use":true,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65535,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://ai.google.dev/pricing/gemini-3","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini/gemini-2.5-pro-exp-03-25":{"mode":"chat","base_model":"gemini-2.5-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":5,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gemini/gemini-2.5-pro-preview-03-25":{"mode":"chat","base_model":"gemini-2.5-pro","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.5-pro-preview-05-06":{"mode":"chat","base_model":"gemini-2.5-pro","deprecation_date":"2025-12-02","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/gemini-2.5-pro-preview-06-05":{"mode":"chat","base_model":"gemini-2.5-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gemini/gemini-2.5-pro-preview-tts":{"mode":"chat","base_model":"gemini-2.5-pro-tts","deprecation_date":"2026-12-18","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-flash-tts","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":false,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":10000000,"supports_audio_input":false,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-exp-1114":{"mode":"chat","base_model":"gemini","max_input_tokens":1048576,"max_output_tokens":8192,"metadata":{"notes":"Rate limits not documented for gemini-exp-1114. Assuming same as gemini-1.5-pro.","supports_tool_choice":true},"rpm":1000,"source":"https://ai.google.dev/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"tpm":4000000,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-exp-1206":{"mode":"chat","base_model":"gemini","max_input_tokens":2097152,"max_output_tokens":8192,"rpm":1000,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"metadata":{"notes":"Rate limits not documented for gemini-exp-1206. Assuming same as gemini-1.5-pro.","supports_tool_choice":true},"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-gemma-2-27b-it":{"mode":"chat","base_model":"gemma-2-27b-it","max_output_tokens":8192,"max_input_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":250000,"rpm":10,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-gemma-2-9b-it":{"mode":"chat","base_model":"gemma-2-9b-it","max_output_tokens":8192,"max_input_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":250000,"rpm":10,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-pro":{"mode":"chat","base_model":"gemini-pro","max_input_tokens":32760,"max_output_tokens":8192,"max_tokens":8192,"rpd":30000,"rpm":360,"source":"https://ai.google.dev/gemini-api/docs/models/gemini","supports_function_calling":true,"supports_tool_choice":true,"tpm":120000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-pro-vision":{"mode":"chat","base_model":"gemini-pro-vision","max_input_tokens":30720,"max_output_tokens":2048,"max_tokens":2048,"rpd":30000,"rpm":360,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":120000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://aistudio.google.com","supports_audio_output":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_vision":true,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/imagen-3.0-fast-generate-001":{"mode":"image_generation","base_model":"imagen-3.0-fast-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/imagen-3.0-generate-001":{"mode":"image_generation","base_model":"imagen-3.0-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/imagen-3.0-generate-002":{"mode":"image_generation","base_model":"imagen-3.0-generate","deprecation_date":"2025-11-10","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/imagen-4.0-fast-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-fast-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/imagen-4.0-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/imagen-4.0-ultra-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-ultra-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/learnlm-1.5-pro-experimental":{"mode":"chat","base_model":"learnlm-1.5-pro","max_input_tokens":32767,"max_output_tokens":8192,"source":"https://aistudio.google.com","supports_audio_output":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/veo-2.0-generate-001":{"mode":"video_generation","base_model":"veo-2.0-generate","max_input_tokens":1024,"max_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/veo-3.0-fast-generate-preview":{"mode":"video_generation","base_model":"veo-3.0-fast-generate","deprecation_date":"2025-11-12","max_input_tokens":1024,"max_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/veo-3.0-generate-preview":{"mode":"video_generation","base_model":"veo-3.0-generate","deprecation_date":"2025-11-12","max_input_tokens":1024,"max_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini/veo-3.1-fast-generate-preview":{"mode":"video_generation","base_model":"veo-3.1-fast-generate","deprecation_date":"2026-10-22","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-omni-1.2-flash","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/veo-3.1-generate-preview":{"mode":"video_generation","base_model":"veo-3.1-generate","deprecation_date":"2026-10-22","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-omni-1.2-flash","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/veo-3.1-fast-generate-001":{"mode":"video_generation","base_model":"veo-3.1-fast-generate","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/veo-3.1-generate-001":{"mode":"video_generation","base_model":"veo-3.1-generate","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/claude-haiku-4.5":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":128000,"max_output_tokens":16000,"successor_model":"github_copilot/claude-haiku-5.5","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"github_copilot","max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/claude-opus-4.5":{"mode":"chat","base_model":"claude-opus-4-5","max_input_tokens":128000,"max_output_tokens":16000,"max_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_computer_use":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_output_config":true},"github_copilot/claude-opus-41":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":80000,"max_output_tokens":16000,"max_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":128000,"max_output_tokens":16000,"max_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_computer_use":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true},"github_copilot/claude-sonnet-4.5":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":128000,"max_output_tokens":16000,"max_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true},"github_copilot/gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_web_search":true,"supports_url_context":true},"github_copilot/gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true,"supports_reasoning":false},"github_copilot/gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":16384,"max_output_tokens":4096,"supports_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-3.5-turbo-0613":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":16384,"max_output_tokens":4096,"supports_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4":{"mode":"chat","base_model":"gpt-4","max_input_tokens":32768,"max_output_tokens":4096,"supports_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4-0613":{"mode":"chat","base_model":"gpt-4","max_input_tokens":32768,"max_output_tokens":4096,"supports_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4-o-preview":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":64000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2026-06-01","deprecation_source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4.1-2025-04-14":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2026-06-01","deprecation_source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-41-copilot":{"mode":"completion","base_model":"gpt-41-copilot","provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":64000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":64000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":64000,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"provider":"github_copilot","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":64000,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","max_input_tokens":64000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","max_input_tokens":64000,"max_output_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"provider":"github_copilot","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-5":{"mode":"chat","base_model":"gpt-5","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"supports_service_tier":true,"supports_custom_tools":true},"github_copilot/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2026-10-19","max_input_tokens":128000,"max_output_tokens":64000,"source":"https://github.blog/changelog/2026-10-19-github-copilot-models-deprecated-on-october-19-2026","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","is_deprecated":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"github_copilot/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"supports_service_tier":true,"supports_custom_tools":true},"github_copilot/gpt-5.1-codex-max":{"mode":"responses","base_model":"gpt-5.1-codex-max","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_web_search":true},"github_copilot/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_custom_tools":true},"github_copilot/text-embedding-3-small":{"mode":"embedding","base_model":"text-embedding-3-small","max_input_tokens":8191,"provider":"github_copilot","max_tokens":8191},"github_copilot/text-embedding-3-small-inference":{"mode":"embedding","base_model":"text-embedding-3-small-inference","max_input_tokens":8191,"provider":"github_copilot","max_tokens":8191},"github_copilot/text-embedding-ada-002":{"mode":"embedding","base_model":"text-embedding-ada-002","max_input_tokens":8191,"provider":"github_copilot","max_tokens":8191},"chatgpt/gpt-5.2-codex":{"mode":"responses","base_model":"gpt-5.2-codex","max_input_tokens":128000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chatgpt/gpt-5.2":{"mode":"responses","base_model":"gpt-5.2","max_input_tokens":128000,"max_output_tokens":64000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"chatgpt","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chatgpt/gpt-5.1-codex-max":{"mode":"responses","base_model":"gpt-5.1-codex-max","max_input_tokens":128000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chatgpt/gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","max_input_tokens":128000,"max_output_tokens":64000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"provider":"chatgpt","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gigachat/GigaChat-2-Lite":{"mode":"chat","base_model":"gigachat-2-lite","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"gigachat","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gigachat/GigaChat-2-Max":{"mode":"chat","base_model":"gigachat-2-max","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_vision":true,"provider":"gigachat","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gigachat/GigaChat-2-Pro":{"mode":"chat","base_model":"gigachat-2-pro","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_vision":true,"provider":"gigachat","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gigachat/Embeddings":{"mode":"embedding","base_model":"embeddings","max_input_tokens":512,"output_vector_size":1024,"provider":"gigachat","max_tokens":512},"gigachat/Embeddings-2":{"mode":"embedding","base_model":"embeddings-2","max_input_tokens":512,"output_vector_size":1024,"provider":"gigachat","max_tokens":512},"gigachat/EmbeddingsGigaR":{"mode":"embedding","base_model":"embeddingsgigar","max_input_tokens":4096,"output_vector_size":2560,"provider":"gigachat","max_tokens":4096},"gmi/anthropic/claude-opus-4.5":{"mode":"chat","base_model":"claude-opus-4-5","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"supports_vision":true,"supports_output_config":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/anthropic/claude-sonnet-4.5":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/anthropic/claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/anthropic/claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/openai/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/openai/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/openai/gpt-5":{"mode":"chat","base_model":"gpt-5","max_input_tokens":409600,"max_output_tokens":32000,"supports_function_calling":true,"provider":"gmi","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/openai/gpt-4o":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":131072,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/openai/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","max_input_tokens":131072,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/deepseek-ai/DeepSeek-V3.2":{"mode":"chat","base_model":"deepseek-v3.2","max_input_tokens":163840,"max_output_tokens":16384,"supports_function_calling":true,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":163840,"max_output_tokens":16384,"supports_function_calling":true,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/google/gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_vision":true,"provider":"gmi","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_web_search":true,"supports_reasoning":false},"gmi/google/gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65536,"supports_function_calling":true,"supports_vision":true,"supports_system_messages":true,"provider":"gmi","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/moonshotai/Kimi-K2-Thinking":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":262144,"max_output_tokens":16384,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/MiniMaxAI/MiniMax-M2.1":{"mode":"chat","base_model":"minimax-m2.1","max_input_tokens":196608,"max_output_tokens":16384,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/Qwen/Qwen3-VL-235B-A22B-Instruct-FP8":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct-fp8","max_input_tokens":262144,"max_output_tokens":16384,"supports_vision":true,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gmi/zai-org/GLM-4.7-FP8":{"mode":"chat","base_model":"glm-4.7-fp8","max_input_tokens":202752,"max_output_tokens":16384,"provider":"gmi","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"google.gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-12b-it.html","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"google.gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-27b-pt.html","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"google.gemma-3-4b-it":{"mode":"chat","base_model":"gemma-3-4b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-4b-it.html","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"google_pse/search":{"mode":"search","base_model":"search","provider":"google_pse","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"global.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"global.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-14","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"global.anthropic.claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"bedrock_converse_supports_strict_tools":false,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-20250514-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1"],"max_tokens":64000,"is_deprecated":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"global.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_parallel_function_calling":true,"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-3.5-turbo-0125":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-3.5-turbo-0301":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":4097,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-3.5-turbo-0613":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":4097,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-3.5-turbo-1106":{"mode":"chat","base_model":"gpt-3.5-turbo","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-3.5-turbo-16k":{"mode":"chat","base_model":"gpt-3.5-turbo-16k","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","max_input_tokens":16385,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-3.5-turbo-16k-0613":{"mode":"chat","base_model":"gpt-3.5-turbo-16k","max_input_tokens":16385,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-3.5-turbo-instruct":{"mode":"completion","base_model":"gpt-3.5-turbo-instruct","deprecation_date":"2026-09-28","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":8192,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","provider":"text-completion-openai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":false,"supports_tool_choice":false,"supports_parallel_function_calling":false,"supports_reasoning":false,"supports_response_schema":false,"supports_prompt_caching":false,"supports_web_search":false,"supports_service_tier":false,"supports_assistant_prefill":false,"supports_system_messages":false},"gpt-3.5-turbo-instruct-0914":{"mode":"completion","base_model":"gpt-3.5-turbo-instruct","max_input_tokens":8192,"max_output_tokens":4097,"provider":"text-completion-openai","max_tokens":4097,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4097},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":false,"supports_tool_choice":false,"supports_parallel_function_calling":false,"supports_reasoning":false,"supports_response_schema":false,"supports_prompt_caching":false,"supports_web_search":false,"supports_service_tier":false,"supports_assistant_prefill":false,"supports_system_messages":false},"gpt-4":{"mode":"chat","base_model":"gpt-4","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":8192,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_parallel_function_calling":true,"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-4-0125-preview":{"mode":"chat","base_model":"gpt-4","deprecation_date":"2026-03-26","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"gpt-4-0314":{"mode":"chat","base_model":"gpt-4","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4-0613":{"mode":"chat","base_model":"gpt-4","deprecation_date":"2025-06-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":8192,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_parallel_function_calling":true,"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-4-1106-preview":{"mode":"chat","base_model":"gpt-4","deprecation_date":"2026-03-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"logitBias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100}},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"topLogprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presencePenalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-4-1106-vision-preview":{"mode":"chat","base_model":"gpt-4-1106-vision","deprecation_date":"2024-12-06","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"is_deprecated":true},"gpt-4-32k":{"mode":"chat","base_model":"gpt-4-32k","max_input_tokens":32768,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4-32k-0314":{"mode":"chat","base_model":"gpt-4-32k","max_input_tokens":32768,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4-32k-0613":{"mode":"chat","base_model":"gpt-4-32k","max_input_tokens":32768,"max_output_tokens":4096,"max_tokens":4096,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4-turbo":{"mode":"chat","base_model":"gpt-4-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-4-turbo-2024-04-09":{"mode":"chat","base_model":"gpt-4-turbo","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-4-turbo-preview":{"mode":"chat","base_model":"gpt-4-turbo","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4-vision-preview":{"mode":"chat","base_model":"gpt-4-vision","deprecation_date":"2024-12-06","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"is_deprecated":true},"gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_assistant_prefill":true},"gpt-4.1-2025-04-14":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_assistant_prefill":true},"gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_assistant_prefill":true},"gpt-4.1-mini-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_assistant_prefill":true},"gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":false,"supports_assistant_prefill":true},"gpt-4.1-nano-2025-04-14":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-luna","max_input_tokens":1047576,"max_output_tokens":32768,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":32768,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":false,"supports_assistant_prefill":true},"gpt-4.5-preview":{"mode":"chat","base_model":"gpt-4.5","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-4.5-preview-2025-02-27":{"mode":"chat","base_model":"gpt-4.5","deprecation_date":"2025-07-14","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"is_deprecated":true},"gpt-4o":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":true,"supports_assistant_prefill":true,"supports_custom_tools":false},"gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":4096,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":true,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-4o-2024-08-06":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":true,"supports_assistant_prefill":true},"gpt-4o-2024-11-20":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":true,"supports_assistant_prefill":true},"gpt-4o-audio-preview":{"mode":"chat","base_model":"gpt-4o-audio","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-4o-audio-preview-2024-10-01":{"mode":"chat","base_model":"gpt-4o-audio","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-4o-audio-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-audio","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-4o-audio-preview-2025-06-03":{"mode":"chat","base_model":"gpt-4o-audio","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-audio":{"mode":"chat","base_model":"gpt-audio","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/realtime","/v1/batch"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-audio-2025-08-28":{"mode":"chat","base_model":"gpt-audio","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/realtime","/v1/batch"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-audio-mini":{"mode":"chat","base_model":"gpt-audio-mini","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/realtime","/v1/batch"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-audio-mini-2025-10-06":{"mode":"chat","base_model":"gpt-audio-mini","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://developers.openai.com/api/docs/models/gpt-audio-mini","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/realtime","/v1/batch"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-audio-mini-2025-12-15":{"mode":"chat","base_model":"gpt-audio-mini","max_input_tokens":128000,"max_output_tokens":16384,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/realtime","/v1/batch"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-audio-1.5","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":true,"supports_assistant_prefill":true,"supports_custom_tools":false},"gpt-4o-mini-2024-07-18":{"mode":"chat","base_model":"gpt-4o-mini","deprecation_date":"2027-01-14","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]},{"id":"prediction","label":"Predicted Output","helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","type":"text","supported_endpoints":["/v1/chat/completions"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_web_search":true,"supports_assistant_prefill":true},"gpt-4o-mini-audio-preview":{"mode":"chat","base_model":"gpt-4o-mini-audio","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-4o-mini-audio-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-audio","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","max_input_tokens":128000,"max_output_tokens":16384,"successor_model":"gpt-audio-1.5","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-realtime-preview":{"mode":"chat","base_model":"gpt-4o-mini-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-mini-realtime","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","max_input_tokens":128000,"max_output_tokens":4096,"successor_model":"gpt-realtime-2.1-mini","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-search-preview":{"mode":"chat","base_model":"gpt-4o-mini-search","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-search-preview-2025-03-11":{"mode":"chat","base_model":"gpt-4o-mini-search","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-4o-mini-search-preview","provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-transcribe":{"mode":"audio_transcription","base_model":"gpt-4o-mini-transcribe","max_input_tokens":16000,"max_output_tokens":2000,"supported_endpoints":["/v1/audio/transcriptions"],"deprecation_date":"2027-02-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-live-transcribe","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-tts":{"mode":"audio_speech","base_model":"gpt-4o-mini-tts","deprecation_date":"2027-01-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text","audio"],"supported_output_modalities":["audio"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-realtime-preview":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-realtime-preview-2024-10-01":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-realtime-preview-2024-12-17":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_prompt_caching":true},"gpt-4o-realtime-preview-2025-06-03":{"mode":"chat","base_model":"gpt-4o-realtime","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-search-preview":{"mode":"chat","base_model":"gpt-4o-search","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-search-preview-2025-03-11":{"mode":"chat","base_model":"gpt-4o-search","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-4o-search-preview","provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-transcribe":{"mode":"audio_transcription","base_model":"gpt-4o-transcribe","max_input_tokens":16000,"max_output_tokens":2000,"supported_endpoints":["/v1/audio/transcriptions"],"deprecation_date":"2027-02-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-live-transcribe","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","deprecation_date":"2026-12-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-image-2.6-sunburst","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","deprecation_date":"2026-12-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-image-2.6-sunburst","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1536/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1536-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1536/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1536-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1536/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1536-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1536/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1536-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1536/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1536-x-1024/gpt-image-1.5":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1536/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1536-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1536/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1536-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1536/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1536-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1536/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1536-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1024-x-1536/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"1536-x-1024/gpt-image-1.5-2025-12-16":{"mode":"image_generation","base_model":"gpt-image-1.5","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"deprecation_date":"2027-04-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-sol","provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.1-2025-11-13":{"mode":"chat","base_model":"gpt-5.1","deprecation_date":"2027-04-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.1-chat-latest":{"mode":"chat","base_model":"gpt-5.1-chat","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.1-chat-latest","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.2-2025-12-11":{"mode":"chat","base_model":"gpt-5.2","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.2-chat-latest":{"mode":"chat","base_model":"gpt-5.2-chat","deprecation_date":"2026-08-10","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.2-chat-latest","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5.2-pro":{"mode":"responses","base_model":"gpt-5.2-pro","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"],"supports_service_tier":true},"gpt-5.2-pro-2025-12-11":{"mode":"responses","base_model":"gpt-5.2-pro","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"],"supports_service_tier":true},"gpt-5-pro":{"mode":"responses","base_model":"gpt-5-pro","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":128000,"max_output_tokens":272000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":272000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":272000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["high"],"supports_service_tier":true},"gpt-5-pro-2025-10-06":{"mode":"responses","base_model":"gpt-5-pro","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":128000,"max_output_tokens":272000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":272000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":272000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["high"],"supports_service_tier":true},"gpt-5-2025-08-07":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-5-chat":{"mode":"chat","base_model":"gpt-5-chat","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.2-chat-latest","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-5-chat-latest":{"mode":"chat","base_model":"gpt-5-chat","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":16384,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5-chat-latest","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":16384,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-5-codex":{"mode":"responses","base_model":"gpt-5-codex","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5-codex","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":128000,"supports_system_messages":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5.1-codex":{"mode":"responses","base_model":"gpt-5.1-codex","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.1-codex","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":128000,"supports_system_messages":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5.1-codex-max":{"mode":"responses","base_model":"gpt-5.1-codex-max","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.1-codex-max","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":128000,"supports_system_messages":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5.1-codex-mini":{"mode":"responses","base_model":"gpt-5.1-codex-mini","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.1-codex-mini","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":128000,"supports_system_messages":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5.2-codex":{"mode":"responses","base_model":"gpt-5.2-codex","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.2-codex","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":128000,"supports_system_messages":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-terra","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-5-mini-2025-08-07":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-terra","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_service_tier":true,"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-5-nano-2025-08-07":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_service_tier":true,"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true},"gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-image-2.5-sunburst","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","deprecation_date":"2026-12-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-image-2.6-sunburst","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-realtime":{"mode":"chat","base_model":"gpt-realtime","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1","max_input_tokens":32000,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":4096,"model_parameters":[{"id":"instructions","label":"Instructions","helpText":"The default system instructions (i.e. system message) prepended to model calls.","type":"text","supported_endpoints":["/v1/responses","/v1/realtime"]},{"id":"voice","label":"Voice","helpText":"Pre-selected voice used when generating the audio","type":"select","default":"alloy","options":[{"label":"Alloy","value":"alloy"},{"label":"Ash","value":"ash"},{"label":"Ballad","value":"ballad"},{"label":"Coral","value":"coral"},{"label":"Echo","value":"echo"},{"label":"Sage","value":"sage"},{"label":"Shimmer","value":"shimmer"},{"label":"Verse","value":"verse"}],"supported_endpoints":["/v1/chat/completions","/v1/realtime","/v1/audio/speech"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0.6,"max":1.2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_response_output_tokens","label":"Max Response Output Tokens","helpText":"Maximum number of output tokens for a single assistant response, inclusive of tool calls.","type":"number","default":510,"supported_endpoints":["/v1/realtime"]},{"id":"input_audio_noise_reduction","label":"Input Audio Noise Reduction","helpText":"Noise reduction applied to audio input, helpful with VAD and model understanding.","type":"select","accesorKey":"type","options":[{"label":"None","value":"none"},{"label":"Near Field","value":"near_field"},{"label":"Far Field","value":"far_field"}],"supported_endpoints":["/v1/realtime"]},{"id":"speed","label":"Speed","helpText":"The speed of the model's spoken response. ","type":"number","default":1,"range":{"min":0.25,"max":1.5,"step":0.05},"supported_endpoints":["/v1/realtime","/v1/audio/speech"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-realtime-mini":{"mode":"chat","base_model":"gpt-realtime-mini","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-realtime-2025-08-28":{"mode":"chat","base_model":"gpt-realtime","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1","max_input_tokens":32000,"max_output_tokens":4096,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":4096,"model_parameters":[{"id":"instructions","label":"Instructions","helpText":"The default system instructions (i.e. system message) prepended to model calls.","type":"text","supported_endpoints":["/v1/responses","/v1/realtime"]},{"id":"voice","label":"Voice","helpText":"Pre-selected voice used when generating the audio","type":"select","default":"alloy","options":[{"label":"Alloy","value":"alloy"},{"label":"Ash","value":"ash"},{"label":"Ballad","value":"ballad"},{"label":"Coral","value":"coral"},{"label":"Echo","value":"echo"},{"label":"Sage","value":"sage"},{"label":"Shimmer","value":"shimmer"},{"label":"Verse","value":"verse"}],"supported_endpoints":["/v1/chat/completions","/v1/realtime","/v1/audio/speech"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0.6,"max":1.2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_response_output_tokens","label":"Max Response Output Tokens","helpText":"Maximum number of output tokens for a single assistant response, inclusive of tool calls.","type":"number","default":510,"supported_endpoints":["/v1/realtime"]},{"id":"input_audio_noise_reduction","label":"Input Audio Noise Reduction","helpText":"Noise reduction applied to audio input, helpful with VAD and model understanding.","type":"select","accesorKey":"type","options":[{"label":"None","value":"none"},{"label":"Near Field","value":"near_field"},{"label":"Far Field","value":"far_field"}],"supported_endpoints":["/v1/realtime"]},{"id":"speed","label":"Speed","helpText":"The speed of the model's spoken response. ","type":"number","default":1,"range":{"min":0.25,"max":1.5,"step":0.05},"supported_endpoints":["/v1/realtime","/v1/audio/speech"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gradient_ai/alibaba-qwen3-32b":{"mode":"chat","base_model":"qwen3-32b","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":131072,"max_output_tokens":40960,"provider":"gradient_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gradient_ai/anthropic-claude-3-opus":{"mode":"chat","base_model":"claude-3-opus","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":1024,"provider":"gradient_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/anthropic-claude-3.5-haiku":{"mode":"chat","base_model":"claude-3-5-haiku","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":1024,"provider":"gradient_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/anthropic-claude-3.5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":1024,"provider":"gradient_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/anthropic-claude-3.7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":1024,"provider":"gradient_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":32768,"max_output_tokens":8000,"provider":"gradient_ai","max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gradient_ai/llama3-8b-instruct":{"mode":"chat","base_model":"llama-3-8b-instruct","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":8192,"max_output_tokens":512,"provider":"gradient_ai","max_tokens":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/llama3.3-70b-instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":128000,"max_output_tokens":2048,"provider":"gradient_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/mistral-nemo-instruct-2407":{"mode":"chat","base_model":"mistral-nemo-instruct","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":128000,"max_output_tokens":512,"provider":"gradient_ai","max_tokens":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/openai-gpt-4o":{"mode":"chat","base_model":"gpt-4o","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":128000,"max_output_tokens":16384,"provider":"gradient_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/openai-gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":128000,"max_output_tokens":16384,"provider":"gradient_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/openai-o3":{"mode":"chat","base_model":"o3","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":100000,"provider":"gradient_ai","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gradient_ai/openai-o3-mini":{"mode":"chat","base_model":"o3-mini","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supports_tool_choice":false,"max_input_tokens":200000,"max_output_tokens":100000,"provider":"gradient_ai","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lemonade/Qwen3-Coder-30B-A3B-Instruct-GGUF":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":32768,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"lemonade","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lemonade/gpt-oss-20b-mxfp4-GGUF":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"lemonade","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lemonade/gpt-oss-120b-mxfp-GGUF":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"lemonade","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lemonade/Gemma-3-4b-it-GGUF":{"mode":"chat","base_model":"gemma-3-4b-it","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"lemonade","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lemonade/Qwen3-4B-Instruct-2507-GGUF":{"mode":"chat","base_model":"qwen3-4b-instruct","max_input_tokens":262144,"max_output_tokens":32768,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"lemonade","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon-nova/nova-micro-v1":{"mode":"chat","base_model":"nova-micro-v1","max_input_tokens":128000,"max_output_tokens":10000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"amazon_nova","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon-nova/nova-lite-v1":{"mode":"chat","base_model":"nova-lite-v1","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"provider":"amazon_nova","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon-nova/nova-premier-v1":{"mode":"chat","base_model":"nova-premier-v1","max_input_tokens":1000000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_response_schema":true,"supports_vision":true,"provider":"amazon_nova","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"amazon-nova/nova-pro-v1":{"mode":"chat","base_model":"nova-pro-v1","max_input_tokens":300000,"max_output_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"provider":"amazon_nova","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"groq/llama-3.1-8b-instant":{"mode":"chat","base_model":"llama-3.1-8b-instant","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_reasoning":false,"supports_vision":false},"groq/llama-3.3-70b-versatile":{"mode":"chat","base_model":"llama-3.3-70b-versatile","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_native_streaming":true,"supports_reasoning":false,"supports_vision":false},"groq/gemma-7b-it":{"mode":"chat","base_model":"gemma-7b-it","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"groq/meta-llama/llama-guard-4-12b":{"mode":"chat","base_model":"llama-guard-4-12b","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_vision":true},"groq/meta-llama/llama-4-maverick-17b-128e-instruct":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct","max_input_tokens":131072,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"groq","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"groq/meta-llama/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":131072,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"groq","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true},"groq/moonshotai/kimi-k2-instruct-0905":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":262144,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"groq/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":32766,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"groq","max_tokens":32766,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32766},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"groq/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"groq","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"groq/playai-tts":{"mode":"audio_speech","base_model":"playai-tts","max_input_tokens":10000,"max_output_tokens":10000,"max_tokens":10000,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"groq/qwen/qwen3-32b":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131000,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"groq/whisper-large-v3":{"mode":"audio_transcription","base_model":"whisper-large-v3","provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"groq/whisper-large-v3-turbo":{"mode":"audio_transcription","base_model":"whisper-large-v3-turbo","provider":"groq","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hd/1024-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hd/1024-x-1792/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hd/1792-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"heroku/claude-3-5-haiku":{"mode":"chat","base_model":"claude-3-5-haiku","supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"max_input_tokens":200000,"max_output_tokens":8192,"provider":"heroku","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"heroku/claude-3-5-sonnet-latest":{"mode":"chat","base_model":"claude-3-5-sonnet","supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"max_input_tokens":200000,"max_output_tokens":8192,"provider":"heroku","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"heroku/claude-3-7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"max_input_tokens":200000,"max_output_tokens":8192,"provider":"heroku","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"heroku/claude-4-sonnet":{"mode":"chat","base_model":"claude-sonnet-4","supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"max_input_tokens":200000,"max_output_tokens":8192,"provider":"heroku","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"high/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/NousResearch/Hermes-3-Llama-3.1-70B":{"mode":"chat","base_model":"hermes-3-llama-3.1-70b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/Qwen/QwQ-32B":{"mode":"chat","base_model":"qwq-32b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"hyperbolic/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-72b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"hyperbolic/Qwen/Qwen2.5-Coder-32B-Instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/Qwen/Qwen3-235B-A22B":{"mode":"chat","base_model":"qwen3-235b-a22b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"hyperbolic/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/deepseek-ai/DeepSeek-V3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Llama-3.2-3B-Instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Meta-Llama-3-70B-Instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Meta-Llama-3.1-405B-Instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Meta-Llama-3.1-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/meta-llama/Meta-Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"hyperbolic/moonshotai/Kimi-K2-Instruct":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"hyperbolic","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"j2-light":{"mode":"completion","base_model":"j2-light","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ai21","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"j2-mid":{"mode":"completion","base_model":"j2-mid","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ai21","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"j2-ultra":{"mode":"completion","base_model":"j2-ultra","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ai21","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-1.5":{"mode":"chat","base_model":"jamba-1.5","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-1.5-large":{"mode":"chat","base_model":"jamba-1.5-large","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-1.5-large@001":{"mode":"chat","base_model":"jamba-1.5-large","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-1.5-mini":{"mode":"chat","base_model":"jamba-1.5-mini","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-1.5-mini@001":{"mode":"chat","base_model":"jamba-1.5-mini","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-large-1.6":{"mode":"chat","base_model":"jamba-large-1.6","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-large-1.7":{"mode":"chat","base_model":"jamba-large-1.7","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-mini-1.6":{"mode":"chat","base_model":"jamba-mini-1.6","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jamba-mini-1.7":{"mode":"chat","base_model":"jamba-mini-1.7","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"ai21","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jina-reranker-v2-base-multilingual":{"mode":"rerank","base_model":"jina-reranker-v2-base-multilingual","max_input_tokens":1024,"max_output_tokens":1024,"source":"https://api.jina.ai/v1/models","provider":"jina_ai","max_document_chunks_per_query":2048,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"jp.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"jp.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"jp.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"lambda_ai/deepseek-llama3.3-70b":{"mode":"chat","base_model":"deepseek-llama3.3-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/deepseek-r1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/deepseek-r1-671b":{"mode":"chat","base_model":"deepseek-r1-671b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lambda_ai/deepseek-v3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/hermes3-405b":{"mode":"chat","base_model":"hermes3-405b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lambda_ai/hermes3-70b":{"mode":"chat","base_model":"hermes3-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lambda_ai/hermes3-8b":{"mode":"chat","base_model":"hermes3-8b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lambda_ai/lfm-40b":{"mode":"chat","base_model":"lfm-40b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/lfm-7b":{"mode":"chat","base_model":"lfm-7b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama-4-maverick-17b-128e-instruct-fp8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":131072,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":16384,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"lambda_ai/llama3.1-405b-instruct-fp8":{"mode":"chat","base_model":"llama-3.1-405b-instruct-fp8","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.1-70b-instruct-fp8":{"mode":"chat","base_model":"llama-3.1-70b-instruct-fp8","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.1-8b-instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.1-nemotron-70b-instruct-fp8":{"mode":"chat","base_model":"llama-3.1-nemotron-70b-instruct-fp8","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.2-11b-vision-instruct":{"mode":"chat","base_model":"llama-3.2-11b-vision-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.2-3b-instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/llama3.3-70b-instruct-fp8":{"mode":"chat","base_model":"llama-3.3-70b-instruct-fp8","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/qwen25-coder-32b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"lambda_ai/qwen3-32b-fp8":{"mode":"chat","base_model":"qwen3-32b-fp8","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"lambda_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-base":{"mode":"completion","base_model":"luminous-base","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-base-control":{"mode":"chat","base_model":"luminous-base-control","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-extended":{"mode":"completion","base_model":"luminous-extended","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-extended-control":{"mode":"chat","base_model":"luminous-extended-control","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-supreme":{"mode":"completion","base_model":"luminous-supreme","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"luminous-supreme-control":{"mode":"chat","base_model":"luminous-supreme-control","max_tokens":2048,"provider":"aleph_alpha","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"max-x-max/50-steps/stability.stable-diffusion-xl-v0":{"mode":"image_generation","base_model":"stable-diffusion-xl-v0","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"max-x-max/max-steps/stability.stable-diffusion-xl-v0":{"mode":"image_generation","base_model":"stable-diffusion-xl-v0","max_input_tokens":77,"max_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1536/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1536-x-1024/gpt-image-1":{"mode":"image_generation","base_model":"gpt-image-1","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1024-x-1536/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"low/1536-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1024-x-1536/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medium/1536-x-1024/gpt-image-1-mini":{"mode":"image_generation","base_model":"gpt-image-1-mini","supported_endpoints":["/v1/images/generations"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medlm-large":{"mode":"chat","base_model":"medlm-large","max_input_tokens":8192,"max_output_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"medlm-medium":{"mode":"chat","base_model":"medlm-medium","max_input_tokens":32768,"max_output_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama2-13b-chat-v1":{"mode":"chat","base_model":"llama-2-13b-chat-v1","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama2-70b-chat-v1":{"mode":"chat","base_model":"llama-2-70b-chat-v1","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama3-1-405b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-405b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama3-1-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-1-70b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-1-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-1-8b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-2-11b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-11b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-2-1b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-1b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-2-3b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-3b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-2-90b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-90b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama3-3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-3-70b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.md","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta.llama4-maverick-17b-instruct-v1:0":{"mode":"chat","base_model":"llama-4-maverick-17b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-maverick-17b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-maverick-17b-instruct.md","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta.llama4-scout-17b-instruct-v1:0":{"mode":"chat","base_model":"llama-4-scout-17b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-scout-17b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-scout-17b-instruct.md","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta_llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":128000,"max_output_tokens":4028,"source":"https://llama.developer.meta.com/docs/models","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"meta_llama","max_tokens":4028,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":4028},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta_llama/Llama-3.3-8B-Instruct":{"mode":"chat","base_model":"llama-3.3-8b-instruct","max_input_tokens":128000,"max_output_tokens":4028,"source":"https://llama.developer.meta.com/docs/models","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"meta_llama","max_tokens":4028,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4028},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"meta_llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":1000000,"max_output_tokens":4028,"source":"https://llama.developer.meta.com/docs/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"meta_llama","max_tokens":4028,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":4028},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta_llama/Llama-4-Scout-17B-16E-Instruct-FP8":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct-fp8","max_input_tokens":10000000,"max_output_tokens":4028,"source":"https://llama.developer.meta.com/docs/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"meta_llama","max_tokens":4028,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4028},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax.minimax-m2":{"mode":"chat","base_model":"minimax-m2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2.html","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/speech-02-hd":{"mode":"audio_speech","base_model":"speech-02-hd","supported_endpoints":["/v1/audio/speech"],"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/speech-02-turbo":{"mode":"audio_speech","base_model":"speech-02-turbo","supported_endpoints":["/v1/audio/speech"],"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/speech-2.6-hd":{"mode":"audio_speech","base_model":"speech-2.6-hd","supported_endpoints":["/v1/audio/speech"],"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/speech-2.6-turbo":{"mode":"audio_speech","base_model":"speech-2.6-turbo","supported_endpoints":["/v1/audio/speech"],"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/MiniMax-M2.1":{"mode":"chat","base_model":"minimax-m2.1","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"max_input_tokens":1000000,"max_output_tokens":8192,"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/MiniMax-M2.1-lightning":{"mode":"chat","base_model":"minimax-m2.1-lightning","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"max_input_tokens":1000000,"max_output_tokens":8192,"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/MiniMax-M2.5":{"mode":"chat","base_model":"minimax-m2.5","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"max_input_tokens":1000000,"max_output_tokens":8192,"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/MiniMax-M2.5-lightning":{"mode":"chat","base_model":"minimax-m2.5-lightning","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"max_input_tokens":1000000,"max_output_tokens":8192,"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"minimax/MiniMax-M2":{"mode":"chat","base_model":"minimax-m2","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"max_input_tokens":200000,"max_output_tokens":8192,"provider":"minimax","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.magistral-small-2509":{"mode":"chat","base_model":"magistral-small","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.ministral-3-14b-instruct":{"mode":"chat","base_model":"ministral-3-14b-instruct","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.ministral-3-3b-instruct":{"mode":"chat","base_model":"ministral-3-3b-instruct","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.ministral-3-8b-instruct":{"mode":"chat","base_model":"ministral-3-8b-instruct","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.mistral-7b-instruct-v0:2":{"mode":"chat","base_model":"mistral-7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"mistral.mistral-large-2402-v1:0":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.md","supports_function_calling":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.mistral-large-2407-v1:0":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8191,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.mistral-large-3-675b-instruct":{"mode":"chat","base_model":"mistral-large-3-675b-instruct","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.mistral-small-2402-v1:0":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-small.md","supports_function_calling":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.mixtral-8x7b-instruct-v0:1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mixtral-8x7b-instruct.md","supports_tool_choice":true,"provider":"bedrock","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.voxtral-mini-3b-2507":{"mode":"chat","base_model":"voxtral-mini-3b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral.voxtral-small-24b-2507":{"mode":"chat","base_model":"voxtral-small-24b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/codestral-2405":{"mode":"chat","base_model":"codestral","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/codestral-2508":{"mode":"chat","base_model":"codestral","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://mistral.ai/news/codestral-25-08","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/codestral-latest":{"mode":"chat","base_model":"codestral","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/models/model-cards/codestral-25-08","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8191,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/codestral-mamba-latest":{"mode":"chat","base_model":"codestral-mamba","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://mistral.ai/technology/","supports_assistant_prefill":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/devstral-medium-2507":{"mode":"chat","base_model":"devstral-medium","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://mistral.ai/news/devstral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/devstral-small-2505":{"mode":"chat","base_model":"devstral-small","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://mistral.ai/news/devstral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/devstral-small-2507":{"mode":"chat","base_model":"devstral-small","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://mistral.ai/news/devstral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/labs-devstral-small-2512":{"mode":"chat","base_model":"labs-devstral-small","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://docs.mistral.ai/models/devstral-small-2-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/devstral-2512":{"mode":"chat","base_model":"devstral","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://mistral.ai/news/devstral-2-vibe-cli","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":true,"supports_prompt_caching":true},"mistral/magistral-medium-2506":{"mode":"chat","base_model":"magistral-medium","max_input_tokens":40000,"max_output_tokens":40000,"max_tokens":40000,"source":"https://mistral.ai/news/magistral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/magistral-medium-2509":{"mode":"chat","base_model":"magistral-medium","max_input_tokens":40000,"max_output_tokens":40000,"max_tokens":40000,"source":"https://mistral.ai/news/magistral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-ocr-latest":{"mode":"ocr","base_model":"mistral-ocr","supported_endpoints":["/v1/ocr"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-ocr-2505-completion":{"mode":"ocr","base_model":"mistral-ocr-2505-completion","supported_endpoints":["/v1/ocr"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/magistral-medium-latest":{"mode":"chat","base_model":"magistral-medium","max_input_tokens":40000,"max_output_tokens":40000,"reasoning_effort_levels":["none","high"],"source":"https://mistral.ai/news/magistral","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":40000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/magistral-small-2506":{"mode":"chat","base_model":"magistral-small","max_input_tokens":40000,"max_output_tokens":40000,"max_tokens":40000,"source":"https://mistral.ai/pricing#api-pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/magistral-small-latest":{"mode":"chat","base_model":"magistral-small","max_input_tokens":40000,"max_output_tokens":40000,"reasoning_effort_levels":["none","high"],"source":"https://mistral.ai/pricing#api-pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":40000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-embed":{"mode":"embedding","base_model":"mistral-embed","max_input_tokens":8192,"provider":"mistral","max_tokens":8192},"mistral/codestral-embed":{"mode":"embedding","base_model":"codestral-embed","max_input_tokens":8192,"provider":"mistral","max_tokens":8192},"mistral/codestral-embed-2505":{"mode":"embedding","base_model":"codestral-embed","max_input_tokens":8192,"provider":"mistral","max_tokens":8192},"mistral/mistral-large-2402":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-large-2407":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":true,"supports_prompt_caching":true},"mistral/mistral-large-2411":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-large-latest":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.mistral.ai/models/mistral-large-3-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-large-3":{"mode":"chat","base_model":"mistral-large-3","max_input_tokens":256000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/models/mistral-large-3-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-medium":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":32000,"max_output_tokens":8191,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-medium-2312":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-medium-2505":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":131072,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"mistral/mistral-medium-latest":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":131072,"max_output_tokens":8191,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-small":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":8191,"reasoning_effort_levels":["none","high"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/mistral-small-latest":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":8191,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-small-4-0-26-03","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","supports_developer_role":false,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"mistral/mistral-tiny":{"mode":"chat","base_model":"mistral-tiny","max_input_tokens":32000,"max_output_tokens":8191,"supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/open-codestral-mamba":{"mode":"chat","base_model":"codestral-mamba","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://mistral.ai/technology/","supports_assistant_prefill":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/open-mistral-7b":{"mode":"chat","base_model":"mistral-7b","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"mistral/open-mistral-nemo":{"mode":"chat","base_model":"mistral-nemo","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://mistral.ai/technology/","supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/open-mistral-nemo-2407":{"mode":"chat","base_model":"mistral-nemo","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://mistral.ai/technology/","supports_assistant_prefill":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/open-mixtral-8x22b":{"mode":"chat","base_model":"mixtral-8x22b","max_input_tokens":65336,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"mistral/open-mixtral-8x7b":{"mode":"chat","base_model":"mixtral-8x7b","max_input_tokens":32000,"max_output_tokens":8191,"max_tokens":8191,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"mistral/pixtral-12b-2409":{"mode":"chat","base_model":"pixtral-12b","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_parallel_function_calling":false},"mistral/pixtral-large-2411":{"mode":"chat","base_model":"pixtral-large","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"mistral/pixtral-large-latest":{"mode":"chat","base_model":"pixtral-large","max_input_tokens":128000,"max_output_tokens":128000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"moonshot.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_audio_input":false,"supports_response_schema":true,"supports_audio_output":false},"moonshot/kimi-k2-0711-preview":{"mode":"chat","base_model":"kimi-k2","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-k2-0905-preview":{"mode":"chat","base_model":"kimi-k2","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-k2-turbo-preview":{"mode":"chat","base_model":"kimi-k2-turbo","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","deprecation_date":"2026-08-31","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"moonshot","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_prompt_caching":true,"supports_web_search":true,"model":"kimi-k2.5","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-latest":{"mode":"chat","base_model":"kimi","alias_of":"moonshot/kimi-k3","provider":"moonshot","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-latest-128k":{"mode":"chat","base_model":"kimi-latest-128k","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-latest-32k":{"mode":"chat","base_model":"kimi-latest-32k","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-latest-8k":{"mode":"chat","base_model":"kimi-latest-8k","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-thinking-preview":{"mode":"chat","base_model":"kimi-thinking","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_vision":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_video_input":false,"supports_vision":false},"moonshot/kimi-k2-thinking-turbo":{"mode":"chat","base_model":"kimi-k2-thinking-turbo","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"source":"https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-128k":{"mode":"chat","base_model":"moonshot-v1-128k","deprecation_date":"2026-08-31","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"moonshot","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-128k-0430":{"mode":"chat","base_model":"moonshot-v1-128k","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-128k-vision-preview":{"mode":"chat","base_model":"moonshot-v1-128k-vision","deprecation_date":"2026-08-31","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-32k":{"mode":"chat","base_model":"moonshot-v1-32k","deprecation_date":"2026-08-31","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"moonshot","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-32k-0430":{"mode":"chat","base_model":"moonshot-v1-32k","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-32k-vision-preview":{"mode":"chat","base_model":"moonshot-v1-32k-vision","deprecation_date":"2026-08-31","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-8k":{"mode":"chat","base_model":"moonshot-v1-8k","deprecation_date":"2026-08-31","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"moonshot","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-8k-0430":{"mode":"chat","base_model":"moonshot-v1-8k","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"moonshot","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-8k-vision-preview":{"mode":"chat","base_model":"moonshot-v1-8k-vision","deprecation_date":"2026-08-31","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"moonshot","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"moonshot/moonshot-v1-auto":{"mode":"chat","base_model":"moonshot-v1-auto","deprecation_date":"2026-08-31","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://platform.moonshot.ai/docs/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"moonshot","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"morph/morph-v3-fast":{"mode":"chat","base_model":"morph-v3-fast","max_input_tokens":16000,"max_output_tokens":16000,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":false,"provider":"morph","max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"morph/morph-v3-large":{"mode":"chat","base_model":"morph-v3-large","max_input_tokens":16000,"max_output_tokens":16000,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":false,"provider":"morph","max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"multimodalembedding":{"mode":"embedding","base_model":"multimodalembedding","deprecation_date":"2027-04-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":2048,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","supported_endpoints":["/v1/embeddings"],"supported_modalities":["text","image","video"],"supports_embedding_image_input":true,"provider":"vertex_ai","max_tokens":2048},"multimodalembedding@001":{"mode":"embedding","base_model":"multimodalembedding","deprecation_date":"2027-04-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":2048,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","supported_endpoints":["/v1/embeddings"],"supported_modalities":["text","image","video"],"supports_embedding_image_input":true,"provider":"vertex_ai","max_tokens":2048},"nscale/Qwen/QwQ-32B":{"mode":"chat","base_model":"qwq-32b","source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"nscale/Qwen/Qwen2.5-Coder-32B-Instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/Qwen/Qwen2.5-Coder-3B-Instruct":{"mode":"chat","base_model":"qwen2.5-coder-3b-instruct","source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/Qwen/Qwen2.5-Coder-7B-Instruct":{"mode":"chat","base_model":"qwen2.5-coder-7b-instruct","source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"nscale/black-forest-labs/FLUX.1-schnell":{"mode":"image_generation","base_model":"flux.1-schnell","source":"https://docs.nscale.com/docs/inference/serverless-models/current#image-models","supported_endpoints":["/v1/images/generations"],"provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","metadata":{"notes":"Pricing listed as $0.75/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-8B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-8b","metadata":{"notes":"Pricing listed as $0.05/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-1.5b","metadata":{"notes":"Pricing listed as $0.18/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-14b","metadata":{"notes":"Pricing listed as $0.14/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-32b","metadata":{"notes":"Pricing listed as $0.30/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-7b","metadata":{"notes":"Pricing listed as $0.40/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/meta-llama/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","metadata":{"notes":"Pricing listed as $0.06/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","metadata":{"notes":"Pricing listed as $0.40/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nscale/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"nscale/mistralai/mixtral-8x22b-instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x22b-instruct","metadata":{"notes":"Pricing listed as $1.20/1M tokens total. Assumed 50/50 split for input/output."},"source":"https://docs.nscale.com/docs/inference/serverless-models/current#chat-models","provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"nscale/stabilityai/stable-diffusion-xl-base-1.0":{"mode":"image_generation","base_model":"stable-diffusion-xl-base-1.0","source":"https://docs.nscale.com/docs/inference/serverless-models/current#image-models","supported_endpoints":["/v1/images/generations"],"provider":"nscale","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-12b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-9b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nvidia-nemotron-nano-3-30b","max_input_tokens":262144,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"o1":{"mode":"chat","base_model":"o1","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o1-2024-12-17":{"mode":"chat","base_model":"o1","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o1-mini":{"mode":"chat","base_model":"o1-mini","max_input_tokens":128000,"max_output_tokens":65536,"max_tokens":65536,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true},"o1-mini-2024-09-12":{"mode":"chat","base_model":"o1-mini","deprecation_date":"2025-10-27","max_input_tokens":128000,"max_output_tokens":65536,"max_tokens":65536,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"is_deprecated":true},"o1-preview":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"o1-preview-2024-09-12":{"mode":"chat","base_model":"o1","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"openai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"o1-pro":{"mode":"responses","base_model":"o1-pro","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high"],"supports_web_search":false,"supports_service_tier":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":false},"o1-pro-2025-03-19":{"mode":"responses","base_model":"o1-pro","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high"],"supports_web_search":false,"supports_service_tier":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":false},"o3":{"mode":"chat","base_model":"o3","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":100000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o3-2025-04-16":{"mode":"chat","base_model":"o3","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":100000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o3-deep-research":{"mode":"responses","base_model":"o3","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/o3-deep-research","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_pdf_input":true,"deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_system_messages":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"o3-deep-research-2025-06-26":{"mode":"responses","base_model":"o3","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/o3-deep-research","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_pdf_input":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_system_messages":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"o3-mini":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o3-mini-2025-01-31":{"mode":"chat","base_model":"o3-mini","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_web_search":false,"supports_service_tier":true,"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o3-pro":{"mode":"responses","base_model":"o3-pro","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":100000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["low","medium","high"],"supports_service_tier":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":false},"o3-pro-2025-06-10":{"mode":"responses","base_model":"o3-pro","deprecation_date":"2026-12-11","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6.1-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["pro"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":100000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high"],"supports_service_tier":false,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":false},"o4-mini":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o4-mini-2025-04-16":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-10-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["low","medium","high","xhigh"],"supports_assistant_prefill":true,"supports_system_messages":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":true},"o4-mini-deep-research":{"mode":"responses","base_model":"o4-mini","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/o4-mini-deep-research","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_pdf_input":true,"deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_system_messages":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"o4-mini-deep-research-2025-06-26":{"mode":"responses","base_model":"o4-mini","deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/o4-mini-deep-research","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_pdf_input":true,"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"max_tokens":100000,"supports_system_messages":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/meta.llama-3.1-405b-instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/meta.llama-3.2-90b-vision-instruct":{"mode":"chat","base_model":"llama-3.2-90b-vision-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"supports_vision":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/meta.llama-3.3-70b-instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/meta.llama-4-maverick-17b-128e-instruct-fp8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":512000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"supports_vision":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/meta.llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":192000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"oci/xai.grok-3":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/xai.grok-3-fast":{"mode":"chat","base_model":"grok-3-fast","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/xai.grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/xai.grok-3-mini-fast":{"mode":"chat","base_model":"grok-3-mini-fast","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/xai.grok-4":{"mode":"chat","base_model":"grok-4","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/cohere.command-latest":{"mode":"chat","base_model":"command","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/cloud/ai/generative-ai/pricing/","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/cohere.command-a-03-2025":{"mode":"chat","base_model":"command-a","max_input_tokens":256000,"max_output_tokens":4000,"source":"https://www.oracle.com/cloud/ai/generative-ai/pricing/","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"oci/cohere.command-plus-latest":{"mode":"chat","base_model":"command-plus","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/cloud/ai/generative-ai/pricing/","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/codegeex4":{"mode":"chat","base_model":"codegeex4","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":false,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/codegemma":{"mode":"completion","base_model":"codegemma","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/codellama":{"mode":"completion","base_model":"codellama","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/deepseek-coder-v2-base":{"mode":"completion","base_model":"deepseek-coder-v2-base","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/deepseek-coder-v2-instruct":{"mode":"chat","base_model":"deepseek-coder-v2-instruct","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/deepseek-coder-v2-lite-base":{"mode":"completion","base_model":"deepseek-coder-v2-lite-base","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/deepseek-coder-v2-lite-instruct":{"mode":"chat","base_model":"deepseek-coder-v2-lite-instruct","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/deepseek-v3.1:671b-cloud":{"mode":"chat","base_model":"deepseek-v3.1-671b","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"provider":"ollama","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/gpt-oss:120b-cloud":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"provider":"ollama","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/gpt-oss:20b-cloud":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"provider":"ollama","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/internlm2_5-20b-chat":{"mode":"chat","base_model":"internlm2-5-20b-chat","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/llama2":{"mode":"chat","base_model":"llama-2","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/llama2-uncensored":{"mode":"completion","base_model":"llama-2-uncensored","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/llama2:13b":{"mode":"chat","base_model":"llama-2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/llama2:70b":{"mode":"chat","base_model":"llama-2-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/llama2:7b":{"mode":"chat","base_model":"llama-2-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/llama3":{"mode":"chat","base_model":"llama-3","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/llama3.1":{"mode":"chat","base_model":"llama-3.1","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/llama3:70b":{"mode":"chat","base_model":"llama-3-70b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/llama3:8b":{"mode":"chat","base_model":"llama-3-8b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/mistral":{"mode":"completion","base_model":"mistral","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/mistral-7B-Instruct-v0.1":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/mistral-7B-Instruct-v0.2":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"provider":"ollama","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/mistral-large-instruct-2407":{"mode":"chat","base_model":"mistral-large-instruct","max_input_tokens":65536,"max_output_tokens":8192,"supports_function_calling":true,"provider":"ollama","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/mixtral-8x22B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x22b-instruct","max_input_tokens":65536,"max_output_tokens":65536,"supports_function_calling":true,"provider":"ollama","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"provider":"ollama","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/orca-mini":{"mode":"completion","base_model":"orca-mini","max_input_tokens":4096,"max_output_tokens":4096,"provider":"ollama","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ollama/qwen3-coder:480b-cloud":{"mode":"chat","base_model":"qwen3-coder-480b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"provider":"ollama","max_tokens":262144,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ollama/vicuna":{"mode":"completion","base_model":"vicuna","max_input_tokens":2048,"max_output_tokens":2048,"provider":"ollama","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"omni-moderation-2024-09-26":{"mode":"moderation","base_model":"omni-moderation","max_input_tokens":32768,"max_output_tokens":0,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"omni-moderation-latest":{"mode":"moderation","base_model":"omni-moderation","max_input_tokens":32768,"max_output_tokens":0,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"omni-moderation-latest-intents":{"mode":"moderation","base_model":"omni-moderation-latest-intents","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"bedrock_apis":["converse"]},"openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openai.gpt-oss-safeguard-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openai.gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/anthropic/claude-3-haiku":{"mode":"chat","base_model":"claude-3-haiku","max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/anthropic/claude-3.5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/anthropic/claude-3.7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/anthropic/claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/anthropic/claude-opus-4.1":{"mode":"chat","base_model":"claude-opus-4-1","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"supports_computer_use":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/anthropic/claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/anthropic/claude-opus-4.5":{"mode":"chat","base_model":"claude-opus-4-5","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"supports_computer_use":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/anthropic/claude-sonnet-4.5":{"mode":"chat","base_model":"claude-sonnet-4-5","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"supports_computer_use":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/anthropic/claude-haiku-4.5":{"mode":"chat","base_model":"claude-haiku-4-5","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"supports_computer_use":true,"tool_use_system_prompt_tokens":346,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/bytedance/ui-tars-1.5-7b":{"mode":"chat","base_model":"ui-tars-1.5-7b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":2048,"max_tokens":2048,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models/bytedance/ui-tars-1.5-7b","supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-chat":{"mode":"chat","base_model":"deepseek-chat","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-chat-v3-0324":{"mode":"chat","base_model":"deepseek-chat-v3","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-chat-v3.1":{"mode":"chat","base_model":"deepseek-chat","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-v3.2":{"mode":"chat","base_model":"deepseek","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-v3.2-exp":{"mode":"chat","base_model":"deepseek-v3.2","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":false,"supports_response_schema":true,"supports_assistant_prefill":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/deepseek/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":65336,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_native_effort":true},"openrouter/deepseek/deepseek-r1-0528":{"mode":"chat","base_model":"deepseek-r1","provider":"openrouter","max_input_tokens":65336,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_assistant_prefill":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/google/gemini-2.0-flash-001":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":30,"max_tokens":8192,"max_video_length":1,"max_videos_per_prompt":10,"supports_audio_output":true,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"openrouter/google/gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"supports_audio_output":true,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":1,"max":20,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/google/gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"supports_audio_output":true,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/google/gemini-3-pro-preview":{"mode":"chat","base_model":"gemini-3-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/google/gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://ai.google.dev/pricing/gemini-3","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_url_context":true,"supports_web_search":true,"tpm":800000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/gryphe/mythomax-l2-13b":{"mode":"chat","base_model":"mythomax-l2-13b","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":8192,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mancer/weaver":{"mode":"chat","base_model":"weaver","provider":"openrouter","max_input_tokens":8000,"max_output_tokens":6000,"max_tokens":8000,"supports_response_schema":true,"supports_prompt_caching":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/meta-llama/llama-3-70b-instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_tokens":8192,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_system_messages":true,"supports_response_schema":true},"openrouter/minimax/minimax-m2":{"mode":"chat","base_model":"minimax-m2","max_input_tokens":204800,"max_output_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_response_schema":true},"openrouter/mistralai/devstral-2512":{"mode":"chat","base_model":"devstral","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_vision":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/ministral-3b-2512":{"mode":"chat","base_model":"ministral-3b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/ministral-8b-2512":{"mode":"chat","base_model":"ministral-8b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/ministral-14b-2512":{"mode":"chat","base_model":"ministral-14b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mistral-large-2512":{"mode":"chat","base_model":"mistral-large","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mistral-7b-instruct":{"mode":"chat","base_model":"mistral-7b-instruct","max_tokens":8192,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mistral-large":{"mode":"chat","base_model":"mistral-large","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":102400,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mistral-small-3.1-24b-instruct":{"mode":"chat","base_model":"mistral-small-3.1-24b-instruct","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":102400,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mistral-small-3.2-24b-instruct":{"mode":"chat","base_model":"mistral-small-3.2-24b-instruct","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/mistralai/mixtral-8x22b-instruct":{"mode":"chat","base_model":"mixtral-8x22b-instruct","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":52428,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/moonshotai/kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/moonshotai/kimi-k2.5","supports_video_input":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","max_tokens":4095,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":4095},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true},"openrouter/openai/gpt-3.5-turbo-16k":{"mode":"chat","base_model":"gpt-3.5-turbo-16k","provider":"openrouter","max_input_tokens":16385,"max_output_tokens":4096,"max_tokens":16383,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16383},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4":{"mode":"chat","base_model":"gpt-4","provider":"openrouter","max_input_tokens":8191,"max_output_tokens":4096,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4o":{"mode":"chat","base_model":"gpt-4o","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_parallel_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_parallel_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5-chat":{"mode":"chat","base_model":"gpt-5-chat","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_reasoning":true,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":false,"supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true},"openrouter/openai/gpt-5-codex":{"mode":"chat","base_model":"gpt-5-codex","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_reasoning":true,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true},"openrouter/openai/gpt-5.2-codex":{"mode":"chat","base_model":"gpt-5.2-codex","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5":{"mode":"chat","base_model":"gpt-5","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5.2-chat":{"mode":"chat","base_model":"gpt-5.2-chat","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-5.2-pro":{"mode":"chat","base_model":"gpt-5.2-pro","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/openai/gpt-oss-120b","supports_parallel_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/openai/gpt-oss-20b","supports_parallel_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/o1":{"mode":"chat","base_model":"o1","max_input_tokens":200000,"max_output_tokens":100000,"max_tokens":100000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/openai/o3-mini":{"mode":"chat","base_model":"o3-mini","max_input_tokens":128000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"openrouter/openai/o3-mini-high":{"mode":"chat","base_model":"o3-mini-high","max_input_tokens":128000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/qwen/qwen-2.5-coder-32b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","provider":"openrouter","max_input_tokens":33792,"max_output_tokens":33792,"max_tokens":33792,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":33792},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/qwen/qwen-vl-plus":{"mode":"chat","base_model":"qwen-vl-plus","max_input_tokens":8192,"max_output_tokens":2048,"max_tokens":2048,"supports_tool_choice":true,"supports_vision":true,"provider":"openrouter","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/qwen/qwen3-coder":{"mode":"chat","base_model":"qwen3-coder","provider":"openrouter","max_input_tokens":262100,"max_output_tokens":262100,"max_tokens":262100,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/qwen/qwen3-coder","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":262100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/qwen/qwen3-235b-a22b-2507":{"mode":"chat","base_model":"qwen3-235b-a22b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/qwen/qwen3-235b-a22b-2507","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/qwen/qwen3-235b-a22b-thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://openrouter.ai/qwen/qwen3-235b-a22b-thinking-2507","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/switchpoint/router":{"mode":"chat","base_model":"router","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://openrouter.ai/switchpoint/router","supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/undi95/remm-slerp-l2-13b":{"mode":"chat","base_model":"remm-slerp-l2-13b","provider":"openrouter","max_input_tokens":6144,"max_output_tokens":5529,"max_tokens":6144,"supports_response_schema":true,"supports_prompt_caching":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":6144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/x-ai/grok-4":{"mode":"chat","base_model":"grok-4","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://openrouter.ai/x-ai/grok-4","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false,"supports_response_schema":true},"openrouter/z-ai/glm-4.6":{"mode":"chat","base_model":"glm-4.6","provider":"openrouter","max_input_tokens":202800,"max_output_tokens":131000,"max_tokens":131000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/z-ai/glm-4.6","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/z-ai/glm-4.6:exacto":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":202800,"max_output_tokens":131000,"max_tokens":131000,"source":"https://openrouter.ai/z-ai/glm-4.6:exacto","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/xiaomi/mimo-v2-flash":{"mode":"chat","base_model":"mimo-v2-flash","max_input_tokens":262144,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"supports_prompt_caching":false,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_response_schema":true},"openrouter/z-ai/glm-4.7":{"mode":"chat","base_model":"glm-4.7","provider":"openrouter","max_input_tokens":202752,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":false,"supports_vision":true,"supports_assistant_prefill":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/z-ai/glm-4.7-flash":{"mode":"chat","base_model":"glm-4.7-flash","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openrouter/minimax/minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","max_input_tokens":204000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":false,"supports_computer_use":false,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_response_schema":true},"ovhcloud/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/deepseek-r1-distill-llama-70b","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ovhcloud/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/llama-3-1-8b-instruct","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Meta-Llama-3_1-70B-Instruct":{"mode":"chat","base_model":"llama-3-1-70b-instruct","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-1-70b-instruct","supports_function_calling":false,"supports_response_schema":false,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Meta-Llama-3_3-70B-Instruct":{"mode":"chat","base_model":"llama-3-3-70b-instruct","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-3-70b-instruct","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Mistral-7B-Instruct-v0.3":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":127000,"max_output_tokens":127000,"source":"https://endpoints.ai.cloud.ovh.net/models/mistral-7b-instruct-v0-3","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":127000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":127000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ovhcloud/Mistral-Nemo-Instruct-2407":{"mode":"chat","base_model":"mistral-nemo-instruct","max_input_tokens":118000,"max_output_tokens":118000,"source":"https://endpoints.ai.cloud.ovh.net/models/mistral-nemo-instruct-2407","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":118000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":118000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Mistral-Small-3.2-24B-Instruct-2506":{"mode":"chat","base_model":"mistral-small-3.2-24b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://endpoints.ai.cloud.ovh.net/models/mistral-small-3-2-24b-instruct-2506","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"ovhcloud","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://endpoints.ai.cloud.ovh.net/models/mixtral-8x7b-instruct-v0-1","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Qwen2.5-Coder-32B-Instruct":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://endpoints.ai.cloud.ovh.net/models/qwen2-5-coder-32b-instruct","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Qwen2.5-VL-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-vl-72b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://endpoints.ai.cloud.ovh.net/models/qwen2-5-vl-72b-instruct","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"ovhcloud","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/Qwen3-32B":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://endpoints.ai.cloud.ovh.net/models/qwen3-32b","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"ovhcloud","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ovhcloud/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/gpt-oss-120b","supports_function_calling":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ovhcloud/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131000,"max_output_tokens":131000,"source":"https://endpoints.ai.cloud.ovh.net/models/gpt-oss-20b","supports_function_calling":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"ovhcloud/llava-v1.6-mistral-7b-hf":{"mode":"chat","base_model":"llava-v1.6-mistral-7b","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://endpoints.ai.cloud.ovh.net/models/llava-next-mistral-7b","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"ovhcloud","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"ovhcloud/mamba-codestral-7B-v0.1":{"mode":"chat","base_model":"mamba-codestral-7b","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://endpoints.ai.cloud.ovh.net/models/mamba-codestral-7b-v0-1","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false,"provider":"ovhcloud","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/chat-bison":{"mode":"chat","base_model":"chat-bison","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/chat-bison-001":{"mode":"chat","base_model":"chat-bison-001","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/text-bison":{"mode":"completion","base_model":"text-bison","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/text-bison-001":{"mode":"completion","base_model":"text-bison-001","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/text-bison-safety-off":{"mode":"completion","base_model":"text-bison-safety-off","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"palm/text-bison-safety-recitation-off":{"mode":"completion","base_model":"text-bison-safety-recitation-off","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"palm","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"parallel_ai/search":{"mode":"search","base_model":"search","provider":"parallel_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"parallel_ai/search-pro":{"mode":"search","base_model":"search-pro","provider":"parallel_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/codellama-34b-instruct":{"mode":"chat","base_model":"codellama-34b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"perplexity","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"perplexity/codellama-70b-instruct":{"mode":"chat","base_model":"codellama-70b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"perplexity","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"perplexity/llama-2-70b-chat":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"perplexity/llama-3.1-70b-instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"provider":"perplexity","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/llama-3.1-8b-instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"provider":"perplexity","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/llama-3.1-sonar-huge-128k-online":{"mode":"chat","base_model":"llama-3.1-sonar-huge-128k-online","deprecation_date":"2025-02-22","max_input_tokens":127072,"max_output_tokens":127072,"max_tokens":127072,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":127072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"perplexity/llama-3.1-sonar-large-128k-chat":{"mode":"chat","base_model":"llama-3.1-sonar-large-128k-chat","deprecation_date":"2025-02-22","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"perplexity/llama-3.1-sonar-large-128k-online":{"mode":"chat","base_model":"llama-3.1-sonar-large-128k-online","deprecation_date":"2025-02-22","max_input_tokens":127072,"max_output_tokens":127072,"max_tokens":127072,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":127072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"perplexity/llama-3.1-sonar-small-128k-chat":{"mode":"chat","base_model":"llama-3.1-sonar-small-128k-chat","deprecation_date":"2025-02-22","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"perplexity/llama-3.1-sonar-small-128k-online":{"mode":"chat","base_model":"llama-3.1-sonar-small-128k-online","deprecation_date":"2025-02-22","max_input_tokens":127072,"max_output_tokens":127072,"max_tokens":127072,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":127072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"perplexity/mistral-7b-instruct":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"perplexity/mixtral-8x7b-instruct":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/pplx-70b-chat":{"mode":"chat","base_model":"pplx-70b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/pplx-70b-online":{"mode":"chat","base_model":"pplx-70b-online","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/pplx-7b-chat":{"mode":"chat","base_model":"pplx-7b-chat","max_input_tokens":8192,"max_output_tokens":8192,"provider":"perplexity","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/pplx-7b-online":{"mode":"chat","base_model":"pplx-7b-online","max_input_tokens":4096,"max_output_tokens":4096,"provider":"perplexity","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/sonar":{"mode":"chat","base_model":"sonar","max_input_tokens":128000,"max_tokens":128000,"supports_web_search":true,"provider":"perplexity","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.2,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.9,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"search_mode","label":"Search Mode","helpText":"Controls search mode: 'academic' prioritizes scholarly sources, 'sec' prioritizes SEC filings, 'web' uses general web search.","type":"select","default":"web","options":[{"label":"Academic","value":"academic"},{"label":"SEC","value":"sec"},{"label":"Web","value":"web"}]},{"id":"search_domain_filter","label":"Search Domain Filter","helpText":"A comma-separated list of domains to limit search results to. Add a '-' at the beginning of a domain to exclude it. Limited to 20 domains.","type":"text"},{"id":"return_images","label":"Return Images","helpText":"Determines whether search results should include images.","type":"boolean","default":false},{"id":"return_videos","label":"Return Videos","helpText":"Determines whether search results should include videos.","type":"boolean","default":false},{"id":"return_related_questions","label":"Return Related Questions","helpText":"Determines whether related questions should be returned.","type":"boolean","default":false},{"id":"search_recency_filter","label":"Search Recency Filter","helpText":"Filters search results based on time (e.g., 'week', 'day').","type":"text"},{"id":"search_after_date_filter","label":"Search After Date Filter","helpText":"Filters search results to only include content published after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"search_before_date_filter","label":"Search Before Date Filter","helpText":"Filters search results to only include content published before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_after_filter","label":"Last Updated After Filter","helpText":"Filters search results to only include content last updated after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_before_filter","label":"Last Updated Before Filter","helpText":"Filters search results to only include content last updated before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"top_k","label":"Top K","helpText":"The number of tokens to keep for top-k filtering. Limits the model to consider only the k most likely next tokens at each step. A value of 0 disables this filter.","type":"number","default":0},{"id":"disable_search","label":"Disable Search","helpText":"When set to true, disables web search completely and the model will only use its training data to respond.","type":"boolean","default":false},{"id":"enable_search_classifier","label":"Enable Search Classifier","helpText":"Enables a classifier that decides if web search is needed based on your query.","type":"boolean","default":false},{"id":"search_context_size","label":"Search Context Size","helpText":"Controls the size of search context used in web search. Options: low, medium, high.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"language_preference","label":"Language Preference","helpText":"Specifies the preferred language for the chat completion response (e.g., English, Korean, Spanish, etc.).","type":"text"}]},"perplexity/sonar-deep-research":{"mode":"chat","base_model":"sonar","max_input_tokens":128000,"max_output_tokens":128000,"supports_reasoning":true,"supports_web_search":true,"provider":"perplexity","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.2,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.9,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"search_mode","label":"Search Mode","helpText":"Controls search mode: 'academic' prioritizes scholarly sources, 'sec' prioritizes SEC filings, 'web' uses general web search.","type":"select","default":"web","options":[{"label":"Academic","value":"academic"},{"label":"SEC","value":"sec"},{"label":"Web","value":"web"}]},{"id":"search_domain_filter","label":"Search Domain Filter","helpText":"A comma-separated list of domains to limit search results to. Add a '-' at the beginning of a domain to exclude it. Limited to 20 domains.","type":"text"},{"id":"return_images","label":"Return Images","helpText":"Determines whether search results should include images.","type":"boolean","default":false},{"id":"return_videos","label":"Return Videos","helpText":"Determines whether search results should include videos.","type":"boolean","default":false},{"id":"return_related_questions","label":"Return Related Questions","helpText":"Determines whether related questions should be returned.","type":"boolean","default":false},{"id":"search_recency_filter","label":"Search Recency Filter","helpText":"Filters search results based on time (e.g., 'week', 'day').","type":"text"},{"id":"search_after_date_filter","label":"Search After Date Filter","helpText":"Filters search results to only include content published after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"search_before_date_filter","label":"Search Before Date Filter","helpText":"Filters search results to only include content published before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_after_filter","label":"Last Updated After Filter","helpText":"Filters search results to only include content last updated after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_before_filter","label":"Last Updated Before Filter","helpText":"Filters search results to only include content last updated before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"top_k","label":"Top K","helpText":"The number of tokens to keep for top-k filtering. Limits the model to consider only the k most likely next tokens at each step. A value of 0 disables this filter.","type":"number","default":0},{"id":"disable_search","label":"Disable Search","helpText":"When set to true, disables web search completely and the model will only use its training data to respond.","type":"boolean","default":false},{"id":"enable_search_classifier","label":"Enable Search Classifier","helpText":"Enables a classifier that decides if web search is needed based on your query.","type":"boolean","default":false},{"id":"search_context_size","label":"Search Context Size","helpText":"Controls the size of search context used in web search. Options: low, medium, high.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Controls how much computational effort the AI dedicates to each query for deep research models. 'low' provides faster, simpler answers with reduced token usage, 'medium' offers a balanced approach, and 'high' delivers deeper, more thorough responses with increased token usage.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/sonar-medium-chat":{"mode":"chat","base_model":"sonar-medium-chat","max_input_tokens":16384,"max_output_tokens":16384,"provider":"perplexity","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/sonar-medium-online":{"mode":"chat","base_model":"sonar-medium-online","max_input_tokens":12000,"max_output_tokens":12000,"provider":"perplexity","max_tokens":12000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":12000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/sonar-pro":{"mode":"chat","base_model":"sonar-pro","max_input_tokens":200000,"max_output_tokens":8000,"supports_web_search":true,"provider":"perplexity","max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.2,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.9,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"search_mode","label":"Search Mode","helpText":"Controls search mode: 'academic' prioritizes scholarly sources, 'sec' prioritizes SEC filings, 'web' uses general web search.","type":"select","default":"web","options":[{"label":"Academic","value":"academic"},{"label":"SEC","value":"sec"},{"label":"Web","value":"web"}]},{"id":"search_domain_filter","label":"Search Domain Filter","helpText":"A comma-separated list of domains to limit search results to. Add a '-' at the beginning of a domain to exclude it. Limited to 20 domains.","type":"text"},{"id":"return_images","label":"Return Images","helpText":"Determines whether search results should include images.","type":"boolean","default":false},{"id":"return_videos","label":"Return Videos","helpText":"Determines whether search results should include videos.","type":"boolean","default":false},{"id":"return_related_questions","label":"Return Related Questions","helpText":"Determines whether related questions should be returned.","type":"boolean","default":false},{"id":"search_recency_filter","label":"Search Recency Filter","helpText":"Filters search results based on time (e.g., 'week', 'day').","type":"text"},{"id":"search_after_date_filter","label":"Search After Date Filter","helpText":"Filters search results to only include content published after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"search_before_date_filter","label":"Search Before Date Filter","helpText":"Filters search results to only include content published before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_after_filter","label":"Last Updated After Filter","helpText":"Filters search results to only include content last updated after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_before_filter","label":"Last Updated Before Filter","helpText":"Filters search results to only include content last updated before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"top_k","label":"Top K","helpText":"The number of tokens to keep for top-k filtering. Limits the model to consider only the k most likely next tokens at each step. A value of 0 disables this filter.","type":"number","default":0},{"id":"disable_search","label":"Disable Search","helpText":"When set to true, disables web search completely and the model will only use its training data to respond.","type":"boolean","default":false},{"id":"enable_search_classifier","label":"Enable Search Classifier","helpText":"Enables a classifier that decides if web search is needed based on your query.","type":"boolean","default":false},{"id":"search_context_size","label":"Search Context Size","helpText":"Controls the size of search context used in web search. Options: low, medium, high.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"language_preference","label":"Language Preference","helpText":"Specifies the preferred language for the chat completion response (e.g., English, Korean, Spanish, etc.).","type":"text"}]},"perplexity/sonar-reasoning":{"mode":"chat","base_model":"sonar","max_input_tokens":128000,"max_output_tokens":128000,"supports_reasoning":true,"supports_web_search":true,"provider":"perplexity","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.2,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.9,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"search_mode","label":"Search Mode","helpText":"Controls search mode: 'academic' prioritizes scholarly sources, 'sec' prioritizes SEC filings, 'web' uses general web search.","type":"select","default":"web","options":[{"label":"Academic","value":"academic"},{"label":"SEC","value":"sec"},{"label":"Web","value":"web"}]},{"id":"search_domain_filter","label":"Search Domain Filter","helpText":"A comma-separated list of domains to limit search results to. Add a '-' at the beginning of a domain to exclude it. Limited to 20 domains.","type":"text"},{"id":"return_images","label":"Return Images","helpText":"Determines whether search results should include images.","type":"boolean","default":false},{"id":"return_videos","label":"Return Videos","helpText":"Determines whether search results should include videos.","type":"boolean","default":false},{"id":"return_related_questions","label":"Return Related Questions","helpText":"Determines whether related questions should be returned.","type":"boolean","default":false},{"id":"search_recency_filter","label":"Search Recency Filter","helpText":"Filters search results based on time (e.g., 'week', 'day').","type":"text"},{"id":"search_after_date_filter","label":"Search After Date Filter","helpText":"Filters search results to only include content published after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"search_before_date_filter","label":"Search Before Date Filter","helpText":"Filters search results to only include content published before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_after_filter","label":"Last Updated After Filter","helpText":"Filters search results to only include content last updated after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_before_filter","label":"Last Updated Before Filter","helpText":"Filters search results to only include content last updated before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"top_k","label":"Top K","helpText":"The number of tokens to keep for top-k filtering. Limits the model to consider only the k most likely next tokens at each step. A value of 0 disables this filter.","type":"number","default":0},{"id":"disable_search","label":"Disable Search","helpText":"When set to true, disables web search completely and the model will only use its training data to respond.","type":"boolean","default":false},{"id":"enable_search_classifier","label":"Enable Search Classifier","helpText":"Enables a classifier that decides if web search is needed based on your query.","type":"boolean","default":false},{"id":"search_context_size","label":"Search Context Size","helpText":"Controls the size of search context used in web search. Options: low, medium, high.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]}]},"perplexity/sonar-reasoning-pro":{"mode":"chat","base_model":"sonar-reasoning-pro","max_input_tokens":128000,"max_output_tokens":128000,"supports_reasoning":true,"supports_web_search":true,"provider":"perplexity","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.2,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.9,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"search_mode","label":"Search Mode","helpText":"Controls search mode: 'academic' prioritizes scholarly sources, 'sec' prioritizes SEC filings, 'web' uses general web search.","type":"select","default":"web","options":[{"label":"Academic","value":"academic"},{"label":"SEC","value":"sec"},{"label":"Web","value":"web"}]},{"id":"search_domain_filter","label":"Search Domain Filter","helpText":"A comma-separated list of domains to limit search results to. Add a '-' at the beginning of a domain to exclude it. Limited to 20 domains.","type":"text"},{"id":"return_images","label":"Return Images","helpText":"Determines whether search results should include images.","type":"boolean","default":false},{"id":"return_videos","label":"Return Videos","helpText":"Determines whether search results should include videos.","type":"boolean","default":false},{"id":"return_related_questions","label":"Return Related Questions","helpText":"Determines whether related questions should be returned.","type":"boolean","default":false},{"id":"search_recency_filter","label":"Search Recency Filter","helpText":"Filters search results based on time (e.g., 'week', 'day').","type":"text"},{"id":"search_after_date_filter","label":"Search After Date Filter","helpText":"Filters search results to only include content published after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"search_before_date_filter","label":"Search Before Date Filter","helpText":"Filters search results to only include content published before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_after_filter","label":"Last Updated After Filter","helpText":"Filters search results to only include content last updated after this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"last_updated_before_filter","label":"Last Updated Before Filter","helpText":"Filters search results to only include content last updated before this date. Format: MM/DD/YYYY (e.g., 3/1/2025).","type":"text"},{"id":"top_k","label":"Top K","helpText":"The number of tokens to keep for top-k filtering. Limits the model to consider only the k most likely next tokens at each step. A value of 0 disables this filter.","type":"number","default":0},{"id":"disable_search","label":"Disable Search","helpText":"When set to true, disables web search completely and the model will only use its training data to respond.","type":"boolean","default":false},{"id":"enable_search_classifier","label":"Enable Search Classifier","helpText":"Enables a classifier that decides if web search is needed based on your query.","type":"boolean","default":false},{"id":"search_context_size","label":"Search Context Size","helpText":"Controls the size of search context used in web search. Options: low, medium, high.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]}]},"perplexity/sonar-small-chat":{"mode":"chat","base_model":"sonar-small-chat","max_input_tokens":16384,"max_output_tokens":16384,"provider":"perplexity","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/sonar-small-online":{"mode":"chat","base_model":"sonar-small-online","max_input_tokens":12000,"max_output_tokens":12000,"provider":"perplexity","max_tokens":12000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":12000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/swiss-ai/apertus-8b-instruct":{"mode":"chat","base_model":"apertus-8b-instruct","max_input_tokens":8192,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/swiss-ai/apertus-70b-instruct":{"mode":"chat","base_model":"apertus-70b-instruct","max_input_tokens":8192,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/aisingapore/Gemma-SEA-LION-v4-27B-IT":{"mode":"chat","base_model":"gemma-sea-lion-v4-27b-it","max_input_tokens":8192,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/BSC-LT/salamandra-7b-instruct-tools-16k":{"mode":"chat","base_model":"salamandra-7b-instruct-tools-16k","max_input_tokens":16384,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/BSC-LT/ALIA-40b-instruct_Q8_0":{"mode":"chat","base_model":"alia-40b-instruct-q8-0","max_input_tokens":8192,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/allenai/Olmo-3-7B-Instruct":{"mode":"chat","base_model":"olmo-3-7b-instruct","max_input_tokens":32768,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/preset/pro-search":{"mode":"responses","base_model":"preset/pro-search","supports_web_search":true,"supports_function_calling":true,"provider":"perplexity","supports_preset":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/openai/gpt-4o":{"mode":"responses","base_model":"gpt-4o","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"perplexity/openai/gpt-4o-mini":{"mode":"responses","base_model":"gpt-4o-mini","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"perplexity/openai/gpt-5.2":{"mode":"responses","base_model":"gpt-5.2","supports_web_search":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/anthropic/claude-3-5-sonnet-20241022":{"mode":"responses","base_model":"claude-3-5-sonnet","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true},"perplexity/anthropic/claude-3-5-haiku-20241022":{"mode":"responses","base_model":"claude-3-5-haiku","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/google/gemini-2.0-flash-exp":{"mode":"responses","base_model":"gemini-2.0-flash","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/google/gemini-2.0-flash-thinking-exp":{"mode":"responses","base_model":"gemini-2.0-flash-thinking","supports_web_search":true,"supports_reasoning":true,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/xai/grok-2-1212":{"mode":"responses","base_model":"grok-2","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"perplexity/xai/grok-2-vision-1212":{"mode":"responses","base_model":"grok-2-vision","supports_web_search":true,"supports_reasoning":false,"provider":"perplexity","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/aisingapore/Qwen-SEA-LION-v4-32B-IT":{"mode":"chat","base_model":"qwen-sea-lion-v4-32b-it","max_input_tokens":32768,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/allenai/Olmo-3-7B-Think":{"mode":"chat","base_model":"olmo-3-7b-think","max_input_tokens":32768,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"publicai/allenai/Olmo-3-32B-Think":{"mode":"chat","base_model":"olmo-3-32b-think","max_input_tokens":32768,"max_output_tokens":4096,"source":"https://platform.publicai.co/docs","supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"provider":"publicai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"qwen.qwen3-coder-480b-a35b-v1:0":{"mode":"chat","base_model":"qwen3-coder-480b-a35b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html","max_input_tokens":262000,"max_output_tokens":65536,"supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_native_structured_output":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html","provider":"bedrock","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"qwen.qwen3-235b-a22b-2507-v1:0":{"mode":"chat","base_model":"qwen3-235b-a22b","max_input_tokens":262144,"max_output_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_native_structured_output":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html","provider":"bedrock","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"qwen.qwen3-coder-30b-a3b-v1:0":{"mode":"chat","base_model":"qwen3-coder-30b-a3b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":262144,"max_output_tokens":131072,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-30b-a3b-instruct.html","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_native_structured_output":true,"provider":"bedrock","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"qwen.qwen3-32b-v1:0":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":131072,"max_output_tokens":16384,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-32b.html","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_native_structured_output":true,"provider":"bedrock","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/index.json","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"qwen.qwen3-vl-235b-a22b":{"mode":"chat","base_model":"qwen3-vl-235b-a22b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_vision":true,"supports_native_structured_output":true,"supports_response_schema":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"recraft/recraftv2":{"mode":"image_generation","base_model":"recraftv2","source":"https://www.recraft.ai/docs#pricing","supported_endpoints":["/v1/images/generations"],"provider":"recraft","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"recraft/recraftv3":{"mode":"image_generation","base_model":"recraftv3","source":"https://www.recraft.ai/docs#pricing","supported_endpoints":["/v1/images/generations"],"provider":"recraft","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-2-13b":{"mode":"chat","base_model":"llama-2-13b","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-13b","supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-2-13b-chat":{"mode":"chat","base_model":"llama-2-13b-chat","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-13b-chat","supports_system_messages":true,"supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/meta/llama-2-70b":{"mode":"chat","base_model":"llama-2-70b","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-70b","supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-2-70b-chat":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-70b-chat","supports_system_messages":true,"supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/meta/llama-2-7b":{"mode":"chat","base_model":"llama-2-7b","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-7b","supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-2-7b-chat":{"mode":"chat","base_model":"llama-2-7b-chat","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/meta/llama-2-7b-chat","supports_system_messages":true,"supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/meta/llama-3-70b":{"mode":"chat","base_model":"llama-3-70b","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"replicate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-3-70b-instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"replicate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/meta/llama-3-8b":{"mode":"chat","base_model":"llama-3-8b","max_input_tokens":8086,"max_output_tokens":8086,"supports_tool_choice":true,"provider":"replicate","max_tokens":8086,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8086},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/meta/llama-3-8b-instruct":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8086,"max_output_tokens":8086,"supports_tool_choice":true,"provider":"replicate","max_tokens":8086,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8086},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/mistralai/mistral-7b-instruct-v0.2":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/mistralai/mistral-7b-v0.1":{"mode":"chat","base_model":"mistral-7b","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://replicate.com/mistralai/mistral-7b-v0.1","supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/mistralai/mixtral-8x7b-instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"supports_tool_choice":true,"provider":"replicate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-5":{"mode":"chat","base_model":"gpt-5","source":"https://replicate.com/openai/gpt-5","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicateopenai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true},"replicate/anthropic/claude-4.5-haiku":{"mode":"chat","base_model":"claude-haiku-4-5","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"prompt_cache_min_tokens":4096,"source":"https://replicate.com/anthropic/claude-4.5-haiku","provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/ibm-granite/granite-3.3-8b-instruct":{"mode":"chat","base_model":"granite-3.3-8b-instruct","source":"https://replicate.com/ibm-granite/granite-3.3-8b-instruct","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-4o":{"mode":"chat","base_model":"gpt-4o","source":"https://replicate.com/openai/gpt-4o","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_audio_input":true,"supports_audio_output":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/o4-mini":{"mode":"chat","base_model":"o4-mini","supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/o1-mini":{"mode":"chat","base_model":"o1-mini","supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/o1":{"mode":"chat","base_model":"o1","supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","source":"https://replicate.com/openai/gpt-4o-mini","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/qwen/qwen3-235b-a22b-instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","source":"https://replicate.com/qwen/qwen3-235b-a22b-instruct-2507","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/anthropic/claude-4-sonnet":{"mode":"chat","base_model":"claude-sonnet-4","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"prompt_cache_min_tokens":1024,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/deepseek-ai/deepseek-v3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":65536,"max_output_tokens":8192,"source":"https://replicate.com/deepseek-ai/deepseek-v3","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/anthropic/claude-3.7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/anthropic/claude-3.5-haiku":{"mode":"chat","base_model":"claude-3-5-haiku","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/anthropic/claude-3.5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/google/gemini-3-pro":{"mode":"chat","base_model":"gemini-3-pro","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"https://replicate.com/google/gemini-3-pro","supports_audio_input":true,"supports_reasoning":true,"supports_video_input":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/anthropic/claude-4.5-sonnet":{"mode":"chat","base_model":"claude-sonnet-4-5","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"prompt_cache_min_tokens":1024,"source":"https://replicate.com/anthropic/claude-4.5-sonnet","provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","source":"https://replicate.com/openai/gpt-4.1","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","source":"https://replicate.com/openai/gpt-4.1-mini","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","source":"https://replicate.com/openai/gpt-5-nano","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","source":"https://replicate.com/openai/gpt-5-mini","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/google/gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","source":"https://replicate.com/google/gemini-2.5-flash","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_image_size":false,"supports_reasoning":true,"supports_video_input":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","source":"https://replicate.com/openai/gpt-oss-120b","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"replicate/deepseek-ai/deepseek-v3.1":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://replicate.com/deepseek-ai/deepseek-v3.1","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/xai/grok-4":{"mode":"chat","base_model":"grok-4","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"replicate/deepseek-ai/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":65536,"max_output_tokens":8192,"source":"https://replicate.com/deepseek-ai/deepseek-r1","supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"rerank-english-v2.0":{"mode":"rerank","base_model":"rerank-english","max_input_tokens":4096,"max_output_tokens":4096,"max_query_tokens":2048,"max_tokens":4096,"provider":"cohere","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"rerank-english-v3.0":{"mode":"rerank","base_model":"rerank-english","max_input_tokens":4096,"max_output_tokens":4096,"provider":"cohere","max_query_tokens":2048,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"rerank-multilingual-v2.0":{"mode":"rerank","base_model":"rerank-multilingual","max_input_tokens":4096,"max_output_tokens":4096,"max_query_tokens":2048,"max_tokens":4096,"provider":"cohere","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"rerank-multilingual-v3.0":{"mode":"rerank","base_model":"rerank-multilingual","max_input_tokens":4096,"max_output_tokens":4096,"provider":"cohere","max_query_tokens":2048,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"rerank-v3.5":{"mode":"rerank","base_model":"rerank","max_input_tokens":4096,"max_output_tokens":4096,"provider":"cohere","max_query_tokens":2048,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia_nim/nvidia/nv-rerankqa-mistral-4b-v3":{"mode":"rerank","base_model":"nv-rerankqa-mistral-4b-v3","provider":"nvidia_nim","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia_nim/nvidia/llama-3_2-nv-rerankqa-1b-v2":{"mode":"rerank","base_model":"llama-3-2-nv-rerankqa-1b-v2","provider":"nvidia_nim","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"nvidia_nim/ranking/nvidia/llama-3.2-nv-rerankqa-1b-v2":{"mode":"rerank","base_model":"ranking/nvidia/llama-3.2-nv-rerankqa-1b-v2","provider":"nvidia_nim","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-13b":{"mode":"completion","base_model":"llama-2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-13b-f":{"mode":"chat","base_model":"llama-2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-70b":{"mode":"completion","base_model":"llama-2-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-70b-b-f":{"mode":"chat","base_model":"llama-2-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-7b":{"mode":"completion","base_model":"llama-2-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sagemaker/meta-textgeneration-llama-2-7b-f":{"mode":"chat","base_model":"llama-2-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"sagemaker","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/DeepSeek-R1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true},"sambanova/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/Llama-4-Maverick-17B-128E-Instruct":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct","max_input_tokens":131072,"max_output_tokens":131072,"metadata":{"notes":"For vision models, images are converted to 6432 input tokens and are billed at that amount"},"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"sambanova","max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"sambanova/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"metadata":{"notes":"For vision models, images are converted to 6432 input tokens and are billed at that amount"},"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_pdf_input":false,"supports_vision":true},"sambanova/Meta-Llama-3.1-405B-Instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"sambanova/Meta-Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"sambanova/Meta-Llama-3.2-1B-Instruct":{"mode":"chat","base_model":"llama-3.2-1b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/Meta-Llama-3.2-3B-Instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/Meta-Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"sambanova","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/Meta-Llama-Guard-3-8B":{"mode":"chat","base_model":"llama-guard-3-8b","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"sambanova/QwQ-32B":{"mode":"chat","base_model":"qwq-32b","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"sambanova/Qwen2-Audio-7B-Instruct":{"mode":"chat","base_model":"qwen2-audio-7b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"source":"https://cloud.sambanova.ai/plans/pricing","supports_audio_input":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sambanova/Qwen3-32B":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"sambanova","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_response_schema":true},"sambanova/DeepSeek-V3.1":{"mode":"chat","base_model":"deepseek-v3.1","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"sambanova/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/claude-3-5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":18000,"max_output_tokens":8192,"supports_computer_use":true,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":32768,"max_output_tokens":8192,"supports_reasoning":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/gemma-7b":{"mode":"chat","base_model":"gemma-7b","max_input_tokens":8000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/jamba-1.5-large":{"mode":"chat","base_model":"jamba-1.5-large","max_input_tokens":256000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/jamba-1.5-mini":{"mode":"chat","base_model":"jamba-1.5-mini","max_input_tokens":256000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/jamba-instruct":{"mode":"chat","base_model":"jamba-instruct","max_input_tokens":256000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/llama2-70b-chat":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":4096,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3-70b":{"mode":"chat","base_model":"llama-3-70b","max_input_tokens":8000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/llama3-8b":{"mode":"chat","base_model":"llama-3-8b","max_input_tokens":8000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/llama3.1-405b":{"mode":"chat","base_model":"llama-3.1-405b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3.1-70b":{"mode":"chat","base_model":"llama-3.1-70b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3.1-8b":{"mode":"chat","base_model":"llama-3.1-8b","max_input_tokens":128000,"max_output_tokens":8192,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3.2-1b":{"mode":"chat","base_model":"llama-3.2-1b","max_input_tokens":128000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3.2-3b":{"mode":"chat","base_model":"llama-3.2-3b","max_input_tokens":128000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/llama3.3-70b":{"mode":"chat","base_model":"llama-3.3-70b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/mistral-7b":{"mode":"chat","base_model":"mistral-7b","max_input_tokens":32000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/mistral-large":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/mistral-large2":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/mixtral-8x7b":{"mode":"chat","base_model":"mixtral-8x7b","max_input_tokens":32000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"snowflake/reka-core":{"mode":"chat","base_model":"reka-core","max_input_tokens":32000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/reka-flash":{"mode":"chat","base_model":"reka-flash","max_input_tokens":100000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/snowflake-arctic":{"mode":"chat","base_model":"snowflake-arctic","max_input_tokens":4096,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/snowflake-llama-3.1-405b":{"mode":"chat","base_model":"snowflake-llama-3.1-405b","max_input_tokens":8000,"max_output_tokens":8192,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"snowflake/snowflake-llama-3.3-70b":{"mode":"chat","base_model":"snowflake-llama-3.3-70b","max_input_tokens":8000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3":{"mode":"image_generation","base_model":"sd3","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3-large":{"mode":"image_generation","base_model":"sd3-large","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3-large-turbo":{"mode":"image_generation","base_model":"sd3-large-turbo","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3-medium":{"mode":"image_generation","base_model":"sd3-medium","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3.5-large":{"mode":"image_generation","base_model":"sd3.5-large","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3.5-large-turbo":{"mode":"image_generation","base_model":"sd3.5-large-turbo","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sd3.5-medium":{"mode":"image_generation","base_model":"sd3.5-medium","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/stable-image-ultra":{"mode":"image_generation","base_model":"stable-image-ultra","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/inpaint":{"mode":"image_edit","base_model":"inpaint","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/outpaint":{"mode":"image_edit","base_model":"outpaint","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/erase":{"mode":"image_edit","base_model":"erase","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/search-and-replace":{"mode":"image_edit","base_model":"search-and-replace","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/search-and-recolor":{"mode":"image_edit","base_model":"search-and-recolor","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/remove-background":{"mode":"image_edit","base_model":"remove-background","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/replace-background-and-relight":{"mode":"image_edit","base_model":"replace-background-and-relight","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/sketch":{"mode":"image_edit","base_model":"sketch","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/structure":{"mode":"image_edit","base_model":"structure","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/style":{"mode":"image_edit","base_model":"style","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/style-transfer":{"mode":"image_edit","base_model":"style-transfer","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/fast":{"mode":"image_edit","base_model":"fast","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/conservative":{"mode":"image_edit","base_model":"conservative","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/creative":{"mode":"image_edit","base_model":"creative","supported_endpoints":["/v1/images/edits"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability/stable-image-core":{"mode":"image_generation","base_model":"stable-image-core","supported_endpoints":["/v1/images/generations"],"provider":"stability","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.sd3-5-large-v1:0":{"mode":"image_generation","base_model":"sd3-5-large","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.sd3-large-v1:0":{"mode":"image_generation","base_model":"sd3-large","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-core-v1:0":{"mode":"image_generation","base_model":"stable-image-core","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-conservative-upscale-v1:0":{"mode":"image_edit","base_model":"stable-conservative-upscale","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-creative-upscale-v1:0":{"mode":"image_edit","base_model":"stable-creative-upscale","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-fast-upscale-v1:0":{"mode":"image_edit","base_model":"stable-fast-upscale","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-outpaint-v1:0":{"mode":"image_edit","base_model":"stable-outpaint","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-control-sketch-v1:0":{"mode":"image_edit","base_model":"stable-image-control-sketch","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-control-structure-v1:0":{"mode":"image_edit","base_model":"stable-image-control-structure","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-erase-object-v1:0":{"mode":"image_edit","base_model":"stable-image-erase-object","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-inpaint-v1:0":{"mode":"image_edit","base_model":"stable-image-inpaint","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-remove-background-v1:0":{"mode":"image_edit","base_model":"stable-image-remove-background","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-search-recolor-v1:0":{"mode":"image_edit","base_model":"stable-image-search-recolor","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-search-replace-v1:0":{"mode":"image_edit","base_model":"stable-image-search-replace","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-style-guide-v1:0":{"mode":"image_edit","base_model":"stable-image-style-guide","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-style-transfer-v1:0":{"mode":"image_edit","base_model":"stable-style-transfer","max_input_tokens":77,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-core-v1:1":{"mode":"image_generation","base_model":"stable-image-core","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-ultra-v1:0":{"mode":"image_generation","base_model":"stable-image-ultra","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"stability.stable-image-ultra-v1:1":{"mode":"image_generation","base_model":"stable-image-ultra","max_input_tokens":77,"provider":"bedrock","max_tokens":77,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":77},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1024-x-1792/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"standard/1792-x-1024/dall-e-3":{"mode":"image_generation","base_model":"dall-e-3","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"linkup/search":{"mode":"search","base_model":"search","provider":"linkup","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"linkup/search-deep":{"mode":"search","base_model":"search-deep","provider":"linkup","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tavily/search":{"mode":"search","base_model":"search","provider":"tavily","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tavily/search-advanced":{"mode":"search","base_model":"search-advanced","provider":"tavily","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-bison":{"mode":"completion","base_model":"text-bison","max_input_tokens":8192,"max_output_tokens":2048,"max_tokens":2048,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-bison32k":{"mode":"completion","base_model":"text-bison32k","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-bison32k@002":{"mode":"completion","base_model":"text-bison32k","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-bison@001":{"mode":"completion","base_model":"text-bison","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-bison@002":{"mode":"completion","base_model":"text-bison","max_input_tokens":8192,"max_output_tokens":1024,"max_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-completion-codestral/codestral-2405":{"mode":"completion","base_model":"codestral","deprecation_date":"2025-06-16","deprecation_source":"https://docs.mistral.ai/getting-started/models/","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/capabilities/code_generation/","provider":"text-completion-codestral","is_deprecated":true,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-completion-codestral/codestral-latest":{"mode":"completion","base_model":"codestral","max_input_tokens":32000,"max_output_tokens":8191,"source":"https://docs.mistral.ai/capabilities/code_generation/","provider":"text-completion-codestral","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-embedding-004":{"mode":"embedding","base_model":"text-embedding-004","deprecation_date":"2026-01-14","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":2048,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","provider":"vertex_ai","is_deprecated":true,"max_tokens":2048},"text-embedding-005":{"mode":"embedding","base_model":"text-embedding-005","deprecation_date":"2027-04-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":2048,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","provider":"vertex_ai","max_tokens":2048},"text-embedding-3-large":{"mode":"embedding","base_model":"text-embedding-3-large","max_input_tokens":8191,"output_vector_size":3072,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":8191},"text-embedding-3-small":{"mode":"embedding","base_model":"text-embedding-3-small","max_input_tokens":8191,"output_vector_size":1536,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":8191},"text-embedding-ada-002":{"mode":"embedding","base_model":"text-embedding-ada-002","max_input_tokens":8191,"output_vector_size":1536,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":8191},"text-embedding-ada-002-v2":{"mode":"embedding","base_model":"text-embedding-ada-002","max_input_tokens":8191,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":8191},"text-embedding-large-exp-03-07":{"mode":"embedding","base_model":"text-embedding-large","max_input_tokens":8192,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","provider":"vertex_ai","max_tokens":8192},"text-embedding-preview-0409":{"mode":"embedding","base_model":"text-embedding","max_input_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","max_tokens":3072},"text-moderation-007":{"mode":"moderation","base_model":"text-moderation-007","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-moderation-latest":{"mode":"moderation","base_model":"text-moderation","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-moderation-stable":{"mode":"moderation","base_model":"text-moderation-stable","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-multilingual-embedding-002":{"mode":"embedding","base_model":"text-multilingual-embedding-002","deprecation_date":"2027-04-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","max_input_tokens":2048,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models","provider":"vertex_ai","max_tokens":2048},"text-multilingual-embedding-preview-0409":{"mode":"embedding","base_model":"text-multilingual-embedding","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"text-unicorn":{"mode":"completion","base_model":"text-unicorn","max_input_tokens":8192,"max_output_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"text-unicorn@001":{"mode":"completion","base_model":"text-unicorn","max_input_tokens":8192,"max_output_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"textembedding-gecko":{"mode":"embedding","base_model":"textembedding-gecko","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"textembedding-gecko-multilingual":{"mode":"embedding","base_model":"textembedding-gecko-multilingual","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"textembedding-gecko-multilingual@001":{"mode":"embedding","base_model":"textembedding-gecko-multilingual","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"textembedding-gecko@001":{"mode":"embedding","base_model":"textembedding-gecko","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"textembedding-gecko@003":{"mode":"embedding","base_model":"textembedding-gecko","max_input_tokens":3072,"max_tokens":3072,"output_vector_size":768,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai"},"together-ai-21.1b-41b":{"mode":"chat","base_model":"together-ai-21.1b-41b","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together-ai-4.1b-8b":{"mode":"chat","base_model":"together-ai-4.1b-8b","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together-ai-41.1b-80b":{"mode":"chat","base_model":"together-ai-41.1b-80b","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together-ai-8.1b-21b":{"mode":"chat","base_model":"together-ai-8.1b-21b","max_output_tokens":1000,"max_input_tokens":1000,"provider":"together_ai","max_tokens":1000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together-ai-81.1b-110b":{"mode":"chat","base_model":"together-ai-81.1b-110b","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together-ai-embedding-151m-to-350m":{"mode":"embedding","base_model":"together-ai-embedding-151m-to-350m","provider":"together_ai"},"together-ai-embedding-up-to-150m":{"mode":"embedding","base_model":"together-ai-embedding-up-to-150m","provider":"together_ai"},"together_ai/baai/bge-base-en-v1.5":{"mode":"embedding","base_model":"bge-base-en","max_input_tokens":512,"output_vector_size":768,"provider":"together_ai"},"together_ai/BAAI/bge-base-en-v1.5":{"mode":"embedding","base_model":"bge-base-en","max_input_tokens":512,"output_vector_size":768,"provider":"together_ai"},"together-ai-up-to-4b":{"mode":"chat","base_model":"together-ai-up-to-4b","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen2.5-72B-Instruct-Turbo":{"mode":"chat","base_model":"qwen2.5-72b-instruct-turbo","deprecation_date":"2026-02-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen2.5-7B-Instruct-Turbo":{"mode":"chat","base_model":"qwen2.5-7b-instruct-turbo","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"max_input_tokens":32768,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen3-235B-A22B-Instruct-2507-tput":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":262000,"source":"https://www.together.ai/models/qwen3-235b-a22b-instruct-2507-fp8","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen3-235B-A22B-Thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","max_input_tokens":256000,"source":"https://www.together.ai/models/qwen3-235b-a22b-thinking-2507","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen3-235B-A22B-fp8-tput":{"mode":"chat","base_model":"qwen3-235b-a22b-fp8","max_input_tokens":40000,"source":"https://www.together.ai/models/qwen3-235b-a22b-fp8-tput","supports_function_calling":false,"supports_parallel_function_calling":false,"supports_tool_choice":false,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct-fp8","deprecation_date":"2026-06-04","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":256000,"max_output_tokens":262144,"source":"https://www.together.ai/models/qwen3-coder-480b-a35b-instruct","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":128000,"max_output_tokens":20480,"max_tokens":20480,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_system_messages":true},"together_ai/deepseek-ai/DeepSeek-R1-0528-tput":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":128000,"source":"https://www.together.ai/models/deepseek-r1-0528-throughput","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/deepseek-ai/DeepSeek-V3":{"mode":"chat","base_model":"deepseek-v3","deprecation_date":"2026-10-30","max_input_tokens":65536,"max_output_tokens":8192,"metadata":{"successor":"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813"},"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"deepseek-v3.1","deprecation_date":"2026-05-14","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.together.ai/models/deepseek-v3-1","provider":"together_ai","is_deprecated":true,"max_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/meta-llama/Llama-3.2-3B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.2-3b-instruct-turbo","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.3-70b-instruct-turbo","deprecation_date":"2026-10-22","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://api.together.ai/v1/models","successor_model":"together_ai/meta-models/Muse-Glimmer-30B","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo-Free":{"mode":"chat","base_model":"llama-3.3-70b-instruct-turbo-free","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","deprecation_date":"2026-02-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.1-405b-instruct-turbo","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.1-70b-instruct-turbo","deprecation_date":"2026-02-25","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo":{"mode":"chat","base_model":"llama-3.1-8b-instruct-turbo","deprecation_date":"2026-03-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/mistralai/Mistral-7B-Instruct-v0.1":{"mode":"chat","base_model":"mistral-7b-instruct","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/mistralai/Mistral-Small-24B-Instruct-2501":{"mode":"chat","base_model":"mistral-small-24b-instruct","deprecation_date":"2026-04-02","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/mistralai/Mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"mixtral-8x7b-instruct","deprecation_date":"2026-04-16","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","type":"select","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/moonshotai/Kimi-K2-Instruct":{"mode":"chat","base_model":"kimi-k2-instruct","deprecation_date":"2026-10-30","metadata":{"successor":"together_ai/moonshotai/Kimi-K3"},"source":"https://www.together.ai/models/kimi-k2-instruct","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128000,"max_output_tokens":131072,"source":"https://www.together.ai/models/gpt-oss-120b","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","deprecation_date":"2026-09-15","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":128000,"max_output_tokens":131072,"source":"https://www.together.ai/models/gpt-oss-20b","successor_model":"together_ai/Qwen/Qwen3.5-9B","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/togethercomputer/CodeLlama-34b-Instruct":{"mode":"chat","base_model":"codellama-34b-instruct","deprecation_date":"2026-10-30","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"together_ai/zai-org/GLM-4.5-Air-FP8":{"mode":"chat","base_model":"glm-4.5-air-fp8","deprecation_date":"2026-04-02","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":128000,"max_output_tokens":131072,"source":"https://www.together.ai/models/glm-4-5-air","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/zai-org/GLM-4.6":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":200000,"max_output_tokens":200000,"metadata":{"successor":"together_ai/zai-org/GLM-5.2"},"source":"https://www.together.ai/models/glm-4-6","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"together_ai","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/zai-org/GLM-4.7":{"mode":"chat","base_model":"glm-4.7","deprecation_date":"2026-04-02","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":200000,"max_output_tokens":200000,"source":"https://www.together.ai/models/glm-4-7","provider":"together_ai","is_deprecated":true,"max_tokens":200000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://www.together.ai/models/kimi-k2-5","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_reasoning":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_prompt_caching":true,"supports_response_schema":true},"together_ai/moonshotai/Kimi-K2-Instruct-0905":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":262144,"source":"https://www.together.ai/models/kimi-k2-0905","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"together_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"together_ai/Qwen/Qwen3-Next-80B-A3B-Instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","deprecation_date":"2026-04-02","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://www.together.ai/models/qwen3-next-80b-a3b-instruct","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"together_ai/Qwen/Qwen3-Next-80B-A3B-Thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","deprecation_date":"2026-02-25","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://www.together.ai/models/qwen3-next-80b-a3b-thinking","provider":"together_ai","is_deprecated":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tts-1":{"mode":"audio_speech","base_model":"tts-1","deprecation_date":"2027-01-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tts-1-hd":{"mode":"audio_speech","base_model":"tts-1-hd","deprecation_date":"2027-01-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aws_polly/standard":{"mode":"audio_speech","base_model":"standard","supported_endpoints":["/v1/audio/speech"],"source":"https://aws.amazon.com/polly/pricing/","provider":"aws_polly","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aws_polly/neural":{"mode":"audio_speech","base_model":"neural","supported_endpoints":["/v1/audio/speech"],"source":"https://aws.amazon.com/polly/pricing/","provider":"aws_polly","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aws_polly/long-form":{"mode":"audio_speech","base_model":"long-form","supported_endpoints":["/v1/audio/speech"],"source":"https://aws.amazon.com/polly/pricing/","provider":"aws_polly","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"aws_polly/generative":{"mode":"audio_speech","base_model":"generative","supported_endpoints":["/v1/audio/speech"],"source":"https://aws.amazon.com/polly/pricing/","provider":"aws_polly","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-micro-v1:0":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-premier-v1:0":{"mode":"chat","base_model":"nova-premier","max_input_tokens":1000000,"max_output_tokens":10000,"max_tokens":10000,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_response_schema":true,"supports_vision":true,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.amazon.nova-pro-v1:0":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"max_tokens":10000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-5-haiku-20241022-v1:0":{"mode":"chat","base_model":"claude-3-5-haiku","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"us.anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":4096,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_cache_point":false,"is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-5-sonnet-20241022-v2:0":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-13","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-opus-20240229-v1:0":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_cache_point":false,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-3-sonnet-20240229-v1:0":{"mode":"chat","base_model":"claude-3-sonnet","deprecation_date":"2026-07-30","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-opus-4-1-20250805-v1:0":{"mode":"chat","base_model":"claude-opus-4-1","deprecation_date":"2027-01-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-1-20250805-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2"],"is_deprecated":true,"max_tokens":32000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"us.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"au.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"us.anthropic.claude-opus-4-20250514-v1:0":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","max_tokens":32000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.anthropic.claude-opus-4-5-20251101-v1:0":{"mode":"chat","base_model":"claude-opus-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":false,"bedrock_output_config_effort_ceiling":"high","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-5-20251101-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-opus-4-5-20251101-v1:0":{"mode":"chat","base_model":"claude-opus-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":false,"bedrock_output_config_effort_ceiling":"high","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-5-20251101-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-opus-4-5-20251101-v1:0":{"mode":"chat","base_model":"claude-opus-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":false,"bedrock_output_config_effort_ceiling":"high","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-5-20251101-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"us.anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-10-14","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"us.anthropic.claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"bedrock_converse_supports_strict_tools":false,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-20250514-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1"],"max_tokens":64000,"is_deprecated":true,"tool_use_system_prompt_tokens":159,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["default"],"supports_adaptive_thinking":false,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"us.deepseek.r1-v1:0":{"mode":"chat","base_model":"deepseek-r1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-r1.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-r1.html","supports_function_calling":false,"supports_reasoning":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"eu.deepseek.v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-1-405b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-405b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-1-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-1-70b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-1-8b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-1-8b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-1-8b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-2-11b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-11b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-2-1b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-1b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-1b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-2-3b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-3b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-2-3b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-2-90b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-2-90b-instruct","deprecation_date":"2026-07-07","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/bedrock/USD/current/bedrock.json","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true,"provider":"bedrock","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama3-3-70b-instruct-v1:0":{"mode":"chat","base_model":"llama-3-3-70b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-3-70b-instruct.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama4-maverick-17b-instruct-v1:0":{"mode":"chat","base_model":"llama-4-maverick-17b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-maverick-17b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-maverick-17b-instruct.md","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.meta.llama4-scout-17b-instruct-v1:0":{"mode":"chat","base_model":"llama-4-scout-17b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-scout-17b-instruct.md","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-4-scout-17b-instruct.md","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"us.mistral.pixtral-large-2502-v1:0":{"mode":"chat","base_model":"pixtral-large","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-pixtral-large.md","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"v0/v0-1.0-md":{"mode":"chat","base_model":"v0-1.0-md","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"v0","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"v0/v0-1.5-lg":{"mode":"chat","base_model":"v0-1.5-lg","max_input_tokens":512000,"max_output_tokens":512000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"v0","max_tokens":512000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"v0/v0-1.5-md":{"mode":"chat","base_model":"v0-1.5-md","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"v0","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/alibaba/qwen-3-14b":{"mode":"chat","base_model":"qwen3-14b","max_input_tokens":40960,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/alibaba/qwen-3-235b":{"mode":"chat","base_model":"qwen3-235b","max_input_tokens":40960,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/alibaba/qwen-3-30b":{"mode":"chat","base_model":"qwen3-30b","max_input_tokens":40960,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/alibaba/qwen-3-32b":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":40960,"max_output_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/alibaba/qwen3-coder":{"mode":"chat","base_model":"qwen3-coder","max_input_tokens":262144,"max_output_tokens":66536,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":66536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":66536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/amazon/nova-lite":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/amazon/nova-micro":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/amazon/nova-pro":{"mode":"chat","base_model":"nova-pro","max_input_tokens":300000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/amazon/titan-embed-text-v2":{"mode":"chat","base_model":"titan-embed-text-v2","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3-haiku":{"mode":"chat","base_model":"claude-3-haiku","max_input_tokens":200000,"max_output_tokens":4096,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3-opus":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3.5-haiku":{"mode":"chat","base_model":"claude-3-5-haiku","max_input_tokens":200000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3.5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3.7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":64000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-4-opus":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-4-sonnet":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":200000,"max_output_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3-5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3-5-sonnet-20241022":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-3-7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-haiku-4.5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"vercel_ai_gateway","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-opus-4.1":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"vercel_ai_gateway","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-opus-4.5":{"mode":"chat","base_model":"claude-opus-4-5","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-opus-4.6":{"mode":"chat","base_model":"claude-opus-4-6","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/anthropic/claude-sonnet-4.5":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/cohere/command-a":{"mode":"chat","base_model":"command-a","max_input_tokens":256000,"max_output_tokens":8000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/cohere/command-r":{"mode":"chat","base_model":"command-r","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/cohere/command-r-plus":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/cohere/embed-v4.0":{"mode":"chat","base_model":"embed","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/deepseek/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":128000,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/deepseek/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/deepseek/deepseek-v3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":128000,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/google/gemini-2.0-flash":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-03-31","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vercel_ai_gateway/google/gemini-2.0-flash-lite":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-03-31","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vercel_ai_gateway/google/gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","max_input_tokens":1000000,"max_output_tokens":65536,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_image_size":false,"supports_reasoning":true,"supports_pdf_input":true,"supports_web_search":true,"supports_prompt_caching":true,"provider":"vercel_ai_gateway","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/google/gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_pdf_input":true,"supports_web_search":true,"supports_prompt_caching":true,"provider":"vercel_ai_gateway","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/google/gemini-embedding-001":{"mode":"embedding","base_model":"gemini-embedding-001","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/google/gemma-2-9b":{"mode":"chat","base_model":"gemma-2-9b","max_input_tokens":8192,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/google/text-embedding-005":{"mode":"embedding","base_model":"text-embedding-005","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/google/text-multilingual-embedding-002":{"mode":"embedding","base_model":"text-multilingual-embedding-002","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/inception/mercury-coder-small":{"mode":"chat","base_model":"mercury-coder-small","max_input_tokens":32000,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3-70b":{"mode":"chat","base_model":"llama-3-70b","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3-8b":{"mode":"chat","base_model":"llama-3-8b","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.1-70b":{"mode":"chat","base_model":"llama-3.1-70b","max_input_tokens":128000,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.1-8b":{"mode":"chat","base_model":"llama-3.1-8b","max_input_tokens":131000,"max_output_tokens":131072,"supports_function_calling":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.2-11b":{"mode":"chat","base_model":"llama-3.2-11b","max_input_tokens":128000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.2-1b":{"mode":"chat","base_model":"llama-3.2-1b","max_input_tokens":128000,"max_output_tokens":8192,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.2-3b":{"mode":"chat","base_model":"llama-3.2-3b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.2-90b":{"mode":"chat","base_model":"llama-3.2-90b","max_input_tokens":128000,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-3.3-70b":{"mode":"chat","base_model":"llama-3.3-70b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-4-maverick":{"mode":"chat","base_model":"llama-4-maverick","max_input_tokens":131072,"max_output_tokens":8192,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/meta/llama-4-scout":{"mode":"chat","base_model":"llama-4-scout","max_input_tokens":131072,"max_output_tokens":8192,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/codestral":{"mode":"chat","base_model":"codestral","max_input_tokens":256000,"max_output_tokens":4000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/codestral-embed":{"mode":"chat","base_model":"codestral-embed","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/devstral-small":{"mode":"chat","base_model":"devstral-small","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/magistral-medium":{"mode":"chat","base_model":"magistral-medium","max_input_tokens":128000,"max_output_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/magistral-small":{"mode":"chat","base_model":"magistral-small","max_input_tokens":128000,"max_output_tokens":64000,"supports_function_calling":true,"provider":"vercel_ai_gateway","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/ministral-3b":{"mode":"chat","base_model":"ministral-3b","max_input_tokens":128000,"max_output_tokens":4000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/ministral-8b":{"mode":"chat","base_model":"ministral-8b","max_input_tokens":128000,"max_output_tokens":4000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/mistral-embed":{"mode":"chat","base_model":"mistral-embed","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":0},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/mistral-large":{"mode":"chat","base_model":"mistral-large","max_input_tokens":32000,"max_output_tokens":4000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/mistral-saba-24b":{"mode":"chat","base_model":"mistral-saba-24b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/mistral-small":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":4000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/mixtral-8x22b-instruct":{"mode":"chat","base_model":"mixtral-8x22b-instruct","max_input_tokens":65536,"max_output_tokens":2048,"supports_function_calling":true,"provider":"vercel_ai_gateway","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/pixtral-12b":{"mode":"chat","base_model":"pixtral-12b","max_input_tokens":128000,"max_output_tokens":4000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/mistral/pixtral-large":{"mode":"chat","base_model":"pixtral-large","max_input_tokens":128000,"max_output_tokens":4000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/moonshotai/kimi-k2":{"mode":"chat","base_model":"kimi-k2","max_input_tokens":131072,"max_output_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/morph/morph-v3-fast":{"mode":"chat","base_model":"morph-v3-fast","max_input_tokens":32768,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/morph/morph-v3-large":{"mode":"chat","base_model":"morph-v3-large","max_input_tokens":32768,"max_output_tokens":16384,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-3.5-turbo":{"mode":"chat","base_model":"gpt-3.5-turbo","max_input_tokens":16385,"max_output_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-3.5-turbo-instruct":{"mode":"chat","base_model":"gpt-3.5-turbo-instruct","max_input_tokens":8192,"max_output_tokens":4096,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4-turbo":{"mode":"chat","base_model":"gpt-4-turbo","max_input_tokens":128000,"max_output_tokens":4096,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","max_input_tokens":1047576,"max_output_tokens":32768,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","max_input_tokens":1047576,"max_output_tokens":32768,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","max_input_tokens":1047576,"max_output_tokens":32768,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4o":{"mode":"chat","base_model":"gpt-4o","max_input_tokens":128000,"max_output_tokens":16384,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/gpt-4o-mini":{"mode":"chat","base_model":"gpt-4o-mini","max_input_tokens":128000,"max_output_tokens":16384,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/o1":{"mode":"chat","base_model":"o1","max_input_tokens":200000,"max_output_tokens":100000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/o3":{"mode":"chat","base_model":"o3","max_input_tokens":200000,"max_output_tokens":100000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/o3-mini":{"mode":"chat","base_model":"o3-mini","max_input_tokens":200000,"max_output_tokens":100000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/o4-mini":{"mode":"chat","base_model":"o4-mini","max_input_tokens":200000,"max_output_tokens":100000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"provider":"vercel_ai_gateway","max_tokens":100000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/openai/text-embedding-3-large":{"mode":"embedding","base_model":"text-embedding-3-large","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/openai/text-embedding-3-small":{"mode":"embedding","base_model":"text-embedding-3-small","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/openai/text-embedding-ada-002":{"mode":"embedding","base_model":"text-embedding-ada-002","max_input_tokens":0,"max_output_tokens":0,"provider":"vercel_ai_gateway","max_tokens":0},"vercel_ai_gateway/perplexity/sonar":{"mode":"chat","base_model":"sonar","max_input_tokens":127000,"max_output_tokens":8000,"provider":"vercel_ai_gateway","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/perplexity/sonar-pro":{"mode":"chat","base_model":"sonar-pro","max_input_tokens":200000,"max_output_tokens":8000,"provider":"vercel_ai_gateway","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/perplexity/sonar-reasoning":{"mode":"chat","base_model":"sonar","max_input_tokens":127000,"max_output_tokens":8000,"provider":"vercel_ai_gateway","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/perplexity/sonar-reasoning-pro":{"mode":"chat","base_model":"sonar-reasoning-pro","max_input_tokens":127000,"max_output_tokens":8000,"provider":"vercel_ai_gateway","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/vercel/v0-1.0-md":{"mode":"chat","base_model":"v0-1.0-md","max_input_tokens":128000,"max_output_tokens":32000,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/vercel/v0-1.5-md":{"mode":"chat","base_model":"v0-1.5-md","max_input_tokens":128000,"max_output_tokens":32768,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-2":{"mode":"chat","base_model":"grok-2","max_input_tokens":131072,"max_output_tokens":4000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-2-vision":{"mode":"chat","base_model":"grok-2-vision","max_input_tokens":32768,"max_output_tokens":32768,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-3":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-3-fast":{"mode":"chat","base_model":"grok-3-fast","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-3-mini-fast":{"mode":"chat","base_model":"grok-3-mini-fast","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/xai/grok-4":{"mode":"chat","base_model":"grok-4","max_input_tokens":256000,"max_output_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/zai/glm-4.5":{"mode":"chat","base_model":"glm-4.5","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/zai/glm-4.5-air":{"mode":"chat","base_model":"glm-4.5-air","max_input_tokens":128000,"max_output_tokens":96000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":96000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vercel_ai_gateway/zai/glm-4.6":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":200000,"max_output_tokens":200000,"source":"https://vercel.com/ai-gateway/models/glm-4.6","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"provider":"vercel_ai_gateway","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/chirp":{"mode":"audio_speech","base_model":"chirp","source":"https://cloud.google.com/text-to-speech/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-haiku":{"mode":"chat","base_model":"claude-3-5-haiku","deprecation_date":"2026-07-05","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-haiku@20241022":{"mode":"chat","base_model":"claude-3-5-haiku","deprecation_date":"2026-07-05","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-haiku-4-5@20251001":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2026-10-15","max_input_tokens":200000,"max_output_tokens":8192,"regional_endpoint_uplift_multiplier":1.1,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/haiku-4-5","supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_native_streaming":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"provider":"vertex_ai","supports_web_search":true,"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-sonnet":{"mode":"chat","base_model":"claude-3-5-sonnet","deprecation_date":"2026-02-19","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-sonnet-v2":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-sonnet-v2@20241022":{"mode":"chat","base_model":"claude-3-5-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-5-sonnet@20240620":{"mode":"chat","base_model":"claude-3-5-sonnet","deprecation_date":"2026-02-19","max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-7-sonnet@20250219":{"mode":"chat","base_model":"claude-3-7-sonnet","deprecation_date":"2025-06-01","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vertex_ai/claude-3-haiku":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-08-23","max_input_tokens":200000,"max_output_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-haiku@20240307":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-08-23","max_input_tokens":200000,"max_output_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-opus":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_response_schema":true},"vertex_ai/claude-3-opus@20240229":{"mode":"chat","base_model":"claude-3-opus","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-sonnet":{"mode":"chat","base_model":"claude-3-sonnet","max_input_tokens":200000,"max_output_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-3-sonnet@20240229":{"mode":"chat","base_model":"claude-3-sonnet","max_input_tokens":200000,"max_output_tokens":4096,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-1":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-1@20250805":{"mode":"chat","base_model":"claude-opus-4-1","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-5":{"mode":"chat","base_model":"claude-opus-4-5","deprecation_date":"2026-11-24","successor_model":"vertex_ai/claude-opus-5-5","max_input_tokens":200000,"max_output_tokens":64000,"regional_endpoint_uplift_multiplier":1.1,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai-anthropic_models","supports_web_search":true,"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-5@20251101":{"mode":"chat","base_model":"claude-opus-4-5","deprecation_date":"2026-11-24","max_input_tokens":200000,"max_output_tokens":64000,"regional_endpoint_uplift_multiplier":1.1,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_streaming":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai-anthropic_models","supports_web_search":true,"max_tokens":64000,"tool_use_system_prompt_tokens":159,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"vertex_ai-anthropic_models","supports_web_search":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-09-29","successor_model":"vertex_ai/claude-sonnet-5-5","max_input_tokens":200000,"max_output_tokens":64000,"regional_endpoint_uplift_multiplier":1.1,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai-anthropic_models","supports_web_search":true,"is_deprecated":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-sonnet-4-5@20250929":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-09-29","max_input_tokens":200000,"max_output_tokens":64000,"regional_endpoint_uplift_multiplier":1.1,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_streaming":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","is_deprecated":true,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4@20250514":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-sonnet-4@20250514":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":159,"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistralai/codestral-2@001":{"mode":"chat","base_model":"codestral-2","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/codestral-2":{"mode":"chat","base_model":"codestral-2","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/codestral-2@001":{"mode":"chat","base_model":"codestral-2","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistralai/codestral-2":{"mode":"chat","base_model":"codestral-2","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/codestral-2501":{"mode":"chat","base_model":"codestral","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/codestral@2405":{"mode":"chat","base_model":"codestral","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/codestral@latest":{"mode":"chat","base_model":"codestral","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/deepseek-ai/deepseek-v3.1-maas":{"mode":"chat","base_model":"deepseek-v3.1","deprecation_date":"2026-10-21","max_input_tokens":163840,"max_output_tokens":32768,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_regions":["us-west2"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/deepseek-ai/deepseek-v3.2-maas":{"mode":"chat","base_model":"deepseek-v3.2","deprecation_date":"2026-10-21","max_input_tokens":163840,"max_output_tokens":32768,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_regions":["us-west2"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/deepseek-ai/deepseek-r1-0528-maas":{"mode":"chat","base_model":"deepseek-r1","deprecation_date":"2026-10-21","max_input_tokens":65336,"max_output_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_regions":["us-central1"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/gemini-2.5-flash-image":{"mode":"image_generation","base_model":"gemini-2.5-flash-image","deprecation_date":"2027-03-15","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/gemini-3.1-flash-lite-image","max_input_tokens":32768,"max_output_tokens":32768,"rpm":100000,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-generation#edit-an-image","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","image"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":false,"tpm":8000000,"supports_image_size":false,"provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_tokens":32768,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/gemini-3-pro-image-preview":{"mode":"image_generation","base_model":"gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"supports_reasoning":false,"source":"https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-pro-image","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"provider":"vertex_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/deep-research-pro-preview-12-2025":{"mode":"image_generation","base_model":"deep-research-pro","max_input_tokens":65536,"max_output_tokens":32768,"source":"https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-pro-image","provider":"vertex_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagegeneration@006":{"mode":"image_generation","base_model":"imagegeneration","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-3.0-fast-generate-001":{"mode":"image_generation","base_model":"imagen-3.0-fast-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-3.0-generate-001":{"mode":"image_generation","base_model":"imagen-3.0-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-3.0-generate-002":{"mode":"image_generation","base_model":"imagen-3.0-generate","deprecation_date":"2025-11-10","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vertex_ai/imagen-3.0-capability-001":{"mode":"image_generation","base_model":"imagen-3.0-capability","source":"https://cloud.google.com/vertex-ai/generative-ai/docs/image/edit-insert-objects","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-4.0-fast-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-fast-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-4.0-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/imagen-4.0-ultra-generate-001":{"mode":"image_generation","base_model":"imagen-4.0-ultra-generate","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/jamba-1.5":{"mode":"chat","base_model":"jamba-1.5","deprecation_date":"2026-02-27","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/jamba-1.5-large":{"mode":"chat","base_model":"jamba-1.5-large","deprecation_date":"2026-02-27","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/jamba-1.5-large@001":{"mode":"chat","base_model":"jamba-1.5-large","deprecation_date":"2026-02-27","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/jamba-1.5-mini":{"mode":"chat","base_model":"jamba-1.5-mini","deprecation_date":"2026-02-27","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/jamba-1.5-mini@001":{"mode":"chat","base_model":"jamba-1.5-mini","deprecation_date":"2026-02-27","max_input_tokens":256000,"max_output_tokens":256000,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-3.1-405b-instruct-maas":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas","supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-3.1-70b-instruct-maas":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"source":"https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas","supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-3.1-8b-instruct-maas":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":128000,"max_output_tokens":2048,"metadata":{"notes":"VertexAI states that The Llama 3.1 API service for llama-3.1-70b-instruct-maas and llama-3.1-8b-instruct-maas are in public preview and at no cost."},"source":"https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas","supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-3.2-90b-vision-instruct-maas":{"mode":"chat","base_model":"llama-3.2-90b-vision-instruct","max_input_tokens":128000,"max_output_tokens":2048,"metadata":{"notes":"VertexAI states that The Llama 3.2 API service is at no cost during public preview, and will be priced as per dollar-per-1M-tokens at GA."},"source":"https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas","supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-4-maverick-17b-128e-instruct-maas":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":1000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/meta/llama-4-maverick-17b-16e-instruct-maas":{"mode":"chat","base_model":"llama-4-maverick-17b-16e-instruct","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-4-scout-17b-128e-instruct-maas":{"mode":"chat","base_model":"llama-4-scout-17b-128e-instruct","max_input_tokens":10000000,"max_output_tokens":10000000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":10000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":10000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama-4-scout-17b-16e-instruct-maas":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":10000000,"max_output_tokens":10000000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_modalities":["text","image"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":10000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":10000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/meta/llama3-405b-instruct-maas":{"mode":"chat","base_model":"llama-3-405b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supports_tool_choice":true,"provider":"vertex_ai","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/meta/llama3-70b-instruct-maas":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supports_tool_choice":true,"provider":"vertex_ai","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/meta/llama3-8b-instruct-maas":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supports_tool_choice":true,"provider":"vertex_ai","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/minimaxai/minimax-m2-maas":{"mode":"chat","base_model":"minimax-m2","deprecation_date":"2026-10-21","max_input_tokens":196608,"max_output_tokens":196608,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":196608,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/moonshotai/kimi-k2-thinking-maas":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2026-10-21","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/zai-org/glm-4.7-maas":{"mode":"chat","base_model":"glm-4.7","deprecation_date":"2026-10-21","successor_model":"vertex_ai/zai-org/glm-5.2-maas","max_input_tokens":200000,"max_output_tokens":128000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/zai-org/glm-5-maas":{"mode":"chat","base_model":"glm-5","deprecation_date":"2026-10-21","successor_model":"vertex_ai/zai-org/glm-5.2-maas","max_input_tokens":200000,"max_output_tokens":128000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#glm-models","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-medium-3":{"mode":"chat","base_model":"mistral-medium-3","max_input_tokens":128000,"max_output_tokens":8191,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-medium-3@001":{"mode":"chat","base_model":"mistral-medium-3","max_input_tokens":128000,"max_output_tokens":8191,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistralai/mistral-medium-3":{"mode":"chat","base_model":"mistral-medium-3","max_input_tokens":128000,"max_output_tokens":8191,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistralai/mistral-medium-3@001":{"mode":"chat","base_model":"mistral-medium-3","max_input_tokens":128000,"max_output_tokens":8191,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-large-2411":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-large@2407":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-large@2411-001":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-large@latest":{"mode":"chat","base_model":"mistral-large","max_input_tokens":128000,"max_output_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-nemo@2407":{"mode":"chat","base_model":"mistral-nemo","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-nemo@latest":{"mode":"chat","base_model":"mistral-nemo","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-small-2503":{"mode":"chat","base_model":"mistral-small","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-small-2503@001":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":8191,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/mistral-ocr-2505":{"mode":"ocr","base_model":"mistral-ocr","supported_endpoints":["/v1/ocr"],"source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/deepseek-ai/deepseek-ocr-maas":{"mode":"ocr","base_model":"deepseek-ocr","deprecation_date":"2026-10-21","source":"https://cloud.google.com/vertex-ai/pricing","supported_regions":["us-central1"],"provider":"vertex_ai","is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/openai/gpt-oss-120b-maas":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas","supports_pdf_input":false,"supports_reasoning":true,"provider":"vertex_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/openai/gpt-oss-20b-maas":{"mode":"chat","base_model":"gpt-oss-20b","deprecation_date":"2026-10-21","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas","supports_pdf_input":false,"supports_reasoning":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/qwen/qwen3-235b-a22b-instruct-2507-maas":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","deprecation_date":"2026-10-21","max_input_tokens":262144,"max_output_tokens":16384,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/qwen/qwen3-coder-480b-a35b-instruct-maas":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","deprecation_date":"2026-10-21","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/qwen/qwen3-next-80b-a3b-instruct-maas":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","deprecation_date":"2026-10-21","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/qwen/qwen3-next-80b-a3b-thinking-maas":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","deprecation_date":"2026-10-21","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"vertex_ai","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-2.0-generate-001":{"mode":"video_generation","base_model":"veo-2.0-generate","deprecation_date":"2026-06-30","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/veo-3.1-generate-001","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.0-fast-generate-preview":{"mode":"video_generation","base_model":"veo-3.0-fast-generate","deprecation_date":"2025-11-12","max_input_tokens":1024,"max_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vertex_ai/veo-3.0-generate-preview":{"mode":"video_generation","base_model":"veo-3.0-generate","deprecation_date":"2025-11-12","max_input_tokens":1024,"max_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"vertex_ai/veo-3.0-fast-generate-001":{"mode":"video_generation","base_model":"veo-3.0-fast-generate","deprecation_date":"2026-06-30","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/veo-3.1-fast-generate-001","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.0-generate-001":{"mode":"video_generation","base_model":"veo-3.0-generate","deprecation_date":"2026-06-30","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/veo-3.1-generate-001","max_input_tokens":1024,"source":"https://ai.google.dev/gemini-api/docs/video","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","is_deprecated":true,"max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.1-generate-preview":{"mode":"video_generation","base_model":"veo-3.1-generate","max_input_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.1-fast-generate-preview":{"mode":"video_generation","base_model":"veo-3.1-fast-generate","max_input_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.1-generate-001":{"mode":"video_generation","base_model":"veo-3.1-generate","max_input_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo","successor_model":"vertex_ai/gemini-omni-1.2-flash","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/veo-3.1-fast-generate-001":{"mode":"video_generation","base_model":"veo-3.1-fast-generate","max_input_tokens":1024,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo","successor_model":"vertex_ai/gemini-omni-1.2-flash","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"vertex_ai","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"voyage/rerank-2":{"mode":"rerank","base_model":"rerank-2","max_input_tokens":16000,"max_output_tokens":16000,"provider":"voyage","max_query_tokens":16000,"max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"voyage/rerank-2-lite":{"mode":"rerank","base_model":"rerank-2-lite","max_input_tokens":8000,"max_output_tokens":8000,"provider":"voyage","max_query_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"voyage/rerank-2.5":{"mode":"rerank","base_model":"rerank-2.5","max_input_tokens":32000,"max_output_tokens":32000,"provider":"voyage","max_query_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"voyage/rerank-2.5-lite":{"mode":"rerank","base_model":"rerank-2.5-lite","max_input_tokens":32000,"max_output_tokens":32000,"provider":"voyage","max_query_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"voyage/voyage-2":{"mode":"embedding","base_model":"voyage-2","max_input_tokens":4000,"provider":"voyage","max_tokens":4000},"voyage/voyage-3":{"mode":"embedding","base_model":"voyage-3","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-3-large":{"mode":"embedding","base_model":"voyage-3-large","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-3-lite":{"mode":"embedding","base_model":"voyage-3-lite","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-3.5":{"mode":"embedding","base_model":"voyage-3.5","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-3.5-lite":{"mode":"embedding","base_model":"voyage-3.5-lite","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-code-2":{"mode":"embedding","base_model":"voyage-code-2","max_input_tokens":16000,"provider":"voyage","max_tokens":16000},"voyage/voyage-code-3":{"mode":"embedding","base_model":"voyage-code-3","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-context-3":{"mode":"embedding","base_model":"voyage-context-3","max_input_tokens":120000,"provider":"voyage","max_tokens":120000},"voyage/voyage-finance-2":{"mode":"embedding","base_model":"voyage-finance-2","max_input_tokens":32000,"provider":"voyage","max_tokens":32000},"voyage/voyage-large-2":{"mode":"embedding","base_model":"voyage-large-2","max_input_tokens":16000,"provider":"voyage","max_tokens":16000},"voyage/voyage-law-2":{"mode":"embedding","base_model":"voyage-law-2","max_input_tokens":16000,"provider":"voyage","max_tokens":16000},"voyage/voyage-lite-01":{"mode":"embedding","base_model":"voyage-lite-01","max_input_tokens":4096,"provider":"voyage","max_tokens":4096},"voyage/voyage-lite-02-instruct":{"mode":"embedding","base_model":"voyage-lite-02-instruct","max_input_tokens":4000,"provider":"voyage","max_tokens":4000},"voyage/voyage-multimodal-3":{"mode":"embedding","base_model":"voyage-multimodal-3","max_input_tokens":32000,"supports_embedding_image_input":true,"provider":"voyage","max_tokens":32000},"wandb/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","supports_reasoning":true,"max_input_tokens":131072,"max_output_tokens":131072,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"wandb/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_reasoning":true,"max_input_tokens":131072,"max_output_tokens":131072,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"wandb/zai-org/GLM-4.5":{"mode":"chat","base_model":"glm-4.5","max_tokens":131072,"max_input_tokens":131072,"max_output_tokens":131072,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true},"wandb/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_tokens":262144,"max_input_tokens":262144,"max_output_tokens":262144,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"wandb/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","max_tokens":262144,"max_input_tokens":262144,"max_output_tokens":262144,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"wandb/Qwen/Qwen3-235B-A22B-Thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","max_tokens":262144,"max_input_tokens":262144,"max_output_tokens":262144,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"wandb/moonshotai/Kimi-K2-Instruct":{"mode":"chat","base_model":"kimi-k2-instruct","max_tokens":128000,"max_input_tokens":128000,"max_output_tokens":128000,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true},"wandb/meta-llama/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","provider":"wandb","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"wandb/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"deepseek-v3.1","supports_reasoning":true,"max_input_tokens":128000,"max_output_tokens":128000,"source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_prompt_caching":true,"provider":"wandb","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_tokens":161000,"max_input_tokens":161000,"max_output_tokens":161000,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":161000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":false,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":false},"wandb/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_tokens":161000,"max_input_tokens":161000,"max_output_tokens":161000,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":161000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true,"supports_function_calling":true,"supports_prompt_caching":true},"wandb/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"wandb/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_tokens":64000,"max_input_tokens":64000,"max_output_tokens":64000,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"wandb/microsoft/Phi-4-mini-instruct":{"mode":"chat","base_model":"phi-4-mini-instruct","max_tokens":128000,"max_input_tokens":128000,"max_output_tokens":128000,"provider":"wandb","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_pdf_input":false},"watsonx/ibm/granite-3-8b-instruct":{"mode":"chat","base_model":"granite-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":1024,"supports_audio_input":false,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"watsonx","max_tokens":1024,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1024},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/mistralai/mistral-large":{"mode":"chat","base_model":"mistral-large","max_input_tokens":131072,"max_output_tokens":16384,"supports_audio_input":false,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"watsonx","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/bigscience/mt0-xxl-13b":{"mode":"chat","base_model":"mt0-xxl-13b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/core42/jais-13b-chat":{"mode":"chat","base_model":"jais-13b-chat","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/google/flan-t5-xl-3b":{"mode":"chat","base_model":"flan-t5-xl-3b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-13b-chat-v2":{"mode":"chat","base_model":"granite-13b-chat-v2","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-13b-instruct-v2":{"mode":"chat","base_model":"granite-13b-instruct-v2","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-3-3-8b-instruct":{"mode":"chat","base_model":"granite-3-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-4-h-small":{"mode":"chat","base_model":"granite-4-h-small","max_input_tokens":20480,"max_output_tokens":20480,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":20480,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20480},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-guardian-3-2-2b":{"mode":"chat","base_model":"granite-guardian-3-2-2b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-guardian-3-3-8b":{"mode":"chat","base_model":"granite-guardian-3-3-8b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-ttm-1024-96-r2":{"mode":"chat","base_model":"granite-ttm-1024-96-r2","max_input_tokens":512,"max_output_tokens":512,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-ttm-1536-96-r2":{"mode":"chat","base_model":"granite-ttm-1536-96-r2","max_input_tokens":512,"max_output_tokens":512,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-ttm-512-96-r2":{"mode":"chat","base_model":"granite-ttm-512-96-r2","max_input_tokens":512,"max_output_tokens":512,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":512,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/ibm/granite-vision-3-2-2b":{"mode":"chat","base_model":"granite-vision-3-2-2b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":true,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-3-2-11b-vision-instruct":{"mode":"chat","base_model":"llama-3-2-11b-vision-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-3-2-1b-instruct":{"mode":"chat","base_model":"llama-3-2-1b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-3-2-3b-instruct":{"mode":"chat","base_model":"llama-3-2-3b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-3-2-90b-vision-instruct":{"mode":"chat","base_model":"llama-3-2-90b-vision-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-3-3-70b-instruct":{"mode":"chat","base_model":"llama-3-3-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-4-maverick-17b":{"mode":"chat","base_model":"llama-4-maverick-17b","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/meta-llama/llama-guard-3-11b-vision":{"mode":"chat","base_model":"llama-guard-3-11b-vision","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":true,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/mistralai/mistral-medium-2505":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/mistralai/mistral-small-2503":{"mode":"chat","base_model":"mistral-small","max_input_tokens":32000,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"provider":"watsonx","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/mistralai/mistral-small-3-1-24b-instruct-2503":{"mode":"chat","base_model":"mistral-small-3-1-24b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/mistralai/pixtral-12b-2409":{"mode":"chat","base_model":"pixtral-12b","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":true,"provider":"watsonx","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"watsonx/sdaia/allam-1-13b-instruct":{"mode":"chat","base_model":"allam-1-13b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"watsonx/whisper-large-v3-turbo":{"mode":"audio_transcription","base_model":"whisper-large-v3-turbo","supported_endpoints":["/v1/audio/transcriptions"],"provider":"watsonx","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"whisper-1":{"mode":"audio_transcription","base_model":"whisper-1","supported_endpoints":["/v1/audio/transcriptions"],"deprecation_date":"2027-02-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-live-transcribe","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-2":{"mode":"chat","base_model":"grok-2","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-2-1212":{"mode":"chat","base_model":"grok-2","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-2-latest":{"mode":"chat","base_model":"grok-2","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-2-vision":{"mode":"chat","base_model":"grok-2-vision","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-2-vision-1212":{"mode":"chat","base_model":"grok-2-vision","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-2-vision-latest":{"mode":"chat","base_model":"grok-2-vision","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-3":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-beta":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-fast-beta":{"mode":"chat","base_model":"grok-3-fast","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-fast-latest":{"mode":"chat","base_model":"grok-3-fast","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-latest":{"mode":"chat","base_model":"grok-3","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini-beta":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini-fast":{"mode":"chat","base_model":"grok-3-mini-fast","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini-fast-beta":{"mode":"chat","base_model":"grok-3-mini-fast","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini-fast-latest":{"mode":"chat","base_model":"grok-3-mini-fast","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-3-mini-latest":{"mode":"chat","base_model":"grok-3-mini","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"source":"https://x.ai/api#pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-4":{"mode":"chat","base_model":"grok-4","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-4-fast-reasoning":{"mode":"chat","base_model":"grok-4-fast","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"xai/grok-4-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-fast-non","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"xai/grok-4-0709":{"mode":"chat","base_model":"grok-4","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-4-latest":{"mode":"chat","base_model":"grok-4","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-4-1-fast":{"mode":"chat","base_model":"grok-4-1-fast","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_audio_input":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-4-1-fast-reasoning":{"mode":"chat","base_model":"grok-4-1-fast","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_audio_input":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_pdf_input":false},"xai/grok-4-1-fast-reasoning-latest":{"mode":"chat","base_model":"grok-4-1-fast","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_audio_input":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-4-1-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-1-fast-non","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning","supports_audio_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":false},"xai/grok-4-1-fast-non-reasoning-latest":{"mode":"chat","base_model":"grok-4-1-fast-non","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning","supports_audio_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-beta":{"mode":"chat","base_model":"grok","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"xai/grok-code-fast":{"mode":"chat","base_model":"grok-code-fast","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"xai","max_tokens":256000,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-code-fast-1":{"mode":"chat","base_model":"grok-code-fast-1","deprecation_date":"2026-05-15","deprecation_source":"https://docs.x.ai/developers/migration/may-15-retirement","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://docs.x.ai/docs/models","successor_model":"xai/grok-build-0.1","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"xai","max_tokens":256000,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"is_deprecated":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-code-fast-1-0825":{"mode":"chat","base_model":"grok-code-fast-1","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"xai","max_tokens":256000,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"xai/grok-vision-beta":{"mode":"chat","base_model":"grok-vision","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"zai.glm-4.7":{"mode":"chat","base_model":"zai.glm-4.7","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"zai.glm-5","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.7":{"mode":"chat","base_model":"glm-4.7","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.6":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5":{"mode":"chat","base_model":"glm-4.5","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5v":{"mode":"chat","base_model":"glm-4.5v","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5-x":{"mode":"chat","base_model":"glm-4.5-x","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5-air":{"mode":"chat","base_model":"glm-4.5-air","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5-airx":{"mode":"chat","base_model":"glm-4.5-airx","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4-32b-0414-128k":{"mode":"chat","base_model":"glm-4-32b-0414-128k","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"zai/glm-4.5-flash":{"mode":"chat","base_model":"glm-4.5-flash","max_input_tokens":128000,"max_output_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/search_api":{"mode":"vector_store","base_model":"search-api","provider":"vertex_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/container":{"mode":"chat","base_model":"container","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/sora-2":{"mode":"video_generation","base_model":"sora-2","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_resolutions":["720x1280","1280x720"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/sora-2-pro":{"mode":"video_generation","base_model":"sora-2-pro","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_resolutions":["720x1280","1280x720"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/sora-2-pro-high-res":{"mode":"video_generation","base_model":"sora-2-pro","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_resolutions":["1024x1792","1792x1024"],"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/sora-2":{"mode":"video_generation","base_model":"sora-2","deprecation_date":"2026-11-02","source":"https://azure.microsoft.com/en-us/products/ai-services/video-generation","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"azure","is_deprecated":true,"supported_resolutions":["720x1280","1280x720"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/sora-2-pro":{"mode":"video_generation","base_model":"sora-2-pro","source":"https://azure.microsoft.com/en-us/products/ai-services/video-generation","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"azure","supported_resolutions":["720x1280","1280x720"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"azure/sora-2-pro-high-res":{"mode":"video_generation","base_model":"sora-2-pro","source":"https://azure.microsoft.com/en-us/products/ai-services/video-generation","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"azure","supported_resolutions":["1024x1792","1792x1024"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/gen4_turbo":{"mode":"video_generation","base_model":"gen4-turbo","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"metadata":{"comment":"5 credits per second @ $0.01 per credit = $0.05 per second"},"provider":"runwayml","supported_resolutions":["1280x720","720x1280"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/gen4_aleph":{"mode":"video_generation","base_model":"gen4-aleph","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_resolutions":["1280x720","720x1280"],"metadata":{"comment":"15 credits per second @ $0.01 per credit = $0.15 per second"},"provider":"runwayml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/gen3a_turbo":{"mode":"video_generation","base_model":"gen3a-turbo","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_resolutions":["1280x720","720x1280"],"metadata":{"comment":"5 credits per second @ $0.01 per credit = $0.05 per second"},"provider":"runwayml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/gen4_image":{"mode":"image_generation","base_model":"gen4-image","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"metadata":{"comment":"5 credits per 720p image or 8 credits per 1080p image @ $0.01 per credit. Using 5 credits ($0.05) as base cost"},"provider":"runwayml","supported_resolutions":["1280x720","1920x1080"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/gen4_image_turbo":{"mode":"image_generation","base_model":"gen4-image-turbo","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"metadata":{"comment":"2 credits per image (any resolution) @ $0.01 per credit = $0.02 per image"},"provider":"runwayml","supported_resolutions":["1280x720","1920x1080"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"runwayml/eleven_multilingual_v2":{"mode":"audio_speech","base_model":"eleven-multilingual-v2","source":"https://docs.dev.runwayml.com/guides/pricing/","metadata":{"comment":"Estimated cost based on standard TTS pricing. RunwayML uses ElevenLabs models."},"provider":"runwayml","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-coder-480b-a35b-instruct":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_reasoning":true,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-kontext-pro":{"mode":"image_generation","base_model":"flux-kontext-pro","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/SSD-1B":{"mode":"image_generation","base_model":"ssd-1b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/chronos-hermes-13b-v2":{"mode":"chat","base_model":"chronos-hermes-13b-v2","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-13b":{"mode":"chat","base_model":"codellama-13b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-13b-instruct":{"mode":"chat","base_model":"codellama-13b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-13b-python":{"mode":"chat","base_model":"codellama-13b-python","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-34b":{"mode":"chat","base_model":"codellama-34b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-34b-instruct":{"mode":"chat","base_model":"codellama-34b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-34b-python":{"mode":"chat","base_model":"codellama-34b-python","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-70b":{"mode":"chat","base_model":"codellama-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-70b-instruct":{"mode":"chat","base_model":"codellama-70b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-70b-python":{"mode":"chat","base_model":"codellama-70b-python","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-7b":{"mode":"chat","base_model":"codellama-7b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-7b-instruct":{"mode":"chat","base_model":"codellama-7b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-llama-7b-python":{"mode":"chat","base_model":"codellama-7b-python","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/code-qwen-1p5-7b":{"mode":"chat","base_model":"code-qwen-1.5-7b","max_input_tokens":65536,"max_output_tokens":65536,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/codegemma-2b":{"mode":"chat","base_model":"codegemma-2b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/codegemma-7b":{"mode":"chat","base_model":"codegemma-7b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-671b-v2-p1":{"mode":"chat","base_model":"cogito-671b-v2-p1","max_input_tokens":163840,"max_output_tokens":163840,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-3b":{"mode":"chat","base_model":"cogito-v1-llama-3b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-70b":{"mode":"chat","base_model":"cogito-v1-llama-70b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-8b":{"mode":"chat","base_model":"cogito-v1-llama-8b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-v1-preview-qwen-14b":{"mode":"chat","base_model":"cogito-v1-qwen-14b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/cogito-v1-preview-qwen-32b":{"mode":"chat","base_model":"cogito-v1-qwen-32b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-kontext-max":{"mode":"image_generation","base_model":"flux-kontext-max","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/dbrx-instruct":{"mode":"chat","base_model":"dbrx-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-1b-base":{"mode":"chat","base_model":"deepseek-coder-1b-base","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-33b-instruct":{"mode":"chat","base_model":"deepseek-coder-33b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-base":{"mode":"chat","base_model":"deepseek-coder-7b-base","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-base-v1p5":{"mode":"chat","base_model":"deepseek-coder-7b-base-v1.5","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-instruct-v1p5":{"mode":"chat","base_model":"deepseek-coder-7b-instruct-v1.5","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-lite-base":{"mode":"chat","base_model":"deepseek-coder-v2-lite-base","max_input_tokens":163840,"max_output_tokens":163840,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-lite-instruct":{"mode":"chat","base_model":"deepseek-coder-v2-lite-instruct","max_input_tokens":163840,"max_output_tokens":163840,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-prover-v2":{"mode":"chat","base_model":"deepseek-prover-v2","max_input_tokens":163840,"max_output_tokens":163840,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-0528-distill-qwen3-8b":{"mode":"chat","base_model":"deepseek-r1-0528-distill-qwen3-8b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/deepseek-r1-distill-llama-70b","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-llama-8b":{"mode":"chat","base_model":"deepseek-r1-distill-llama-8b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-14b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-14b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-1p5b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-1.5b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-32b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-7b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-7b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v2-lite-chat":{"mode":"chat","base_model":"deepseek-v2-lite-chat","max_input_tokens":163840,"max_output_tokens":163840,"provider":"fireworks_ai","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v2p5":{"mode":"chat","base_model":"deepseek-v2.5","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/devstral-small-2505":{"mode":"chat","base_model":"devstral-small","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/dobby-mini-unhinged-plus-llama-3-1-8b":{"mode":"chat","base_model":"dobby-mini-unhinged-plus-llama-3-1-8b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/dobby-unhinged-llama-3-3-70b-new":{"mode":"chat","base_model":"dobby-unhinged-llama-3-3-70b-new","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/dolphin-2-9-2-qwen2-72b":{"mode":"chat","base_model":"dolphin-2-9-2-qwen2-72b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/dolphin-2p6-mixtral-8x7b":{"mode":"chat","base_model":"dolphin-2.6-mixtral-8x7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/ernie-4p5-21b-a3b-pt":{"mode":"chat","base_model":"ernie-4.5-21b-a3b-pt","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/ernie-4p5-300b-a47b-pt":{"mode":"chat","base_model":"ernie-4.5-300b-a47b-pt","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/fare-20b":{"mode":"chat","base_model":"fare-20b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/firefunction-v1":{"mode":"chat","base_model":"firefunction-v1","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/firellava-13b":{"mode":"chat","base_model":"firellava-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/firesearch-ocr-v6":{"mode":"image_generation","base_model":"fireworks/accounts/fireworks/models/firesearch-ocr-v6","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/fireworks-asr-large":{"mode":"audio_transcription","base_model":"fireworks-asr-large","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/fireworks-asr-v2":{"mode":"audio_transcription","base_model":"fireworks-asr-v2","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-1-dev":{"mode":"chat","base_model":"flux-1-dev","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-1-dev-controlnet-union":{"mode":"chat","base_model":"flux-1-dev-controlnet-union","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-1-dev-fp8":{"mode":"image_generation","base_model":"flux-1-dev-fp8","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-1-schnell":{"mode":"chat","base_model":"flux-1-schnell","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/flux-1-schnell-fp8":{"mode":"image_generation","base_model":"flux-1-schnell-fp8","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gemma-2b-it":{"mode":"chat","base_model":"gemma-2b-it","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gemma-7b":{"mode":"chat","base_model":"gemma-7b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gemma-7b-it":{"mode":"chat","base_model":"gemma-7b-it","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gemma2-9b-it":{"mode":"chat","base_model":"gemma-2-9b-it","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/glm-4p5v":{"mode":"chat","base_model":"glm-4.5v","max_input_tokens":131072,"max_output_tokens":131072,"supports_reasoning":true,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gpt-oss-safeguard-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/hermes-2-pro-mistral-7b":{"mode":"chat","base_model":"hermes-2-pro-mistral-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/internvl3-38b":{"mode":"chat","base_model":"internvl3-38b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/internvl3-78b":{"mode":"chat","base_model":"internvl3-78b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/internvl3-8b":{"mode":"chat","base_model":"internvl3-8b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/japanese-stable-diffusion-xl":{"mode":"image_generation","base_model":"japanese-stable-diffusion-xl","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kat-coder":{"mode":"chat","base_model":"kat-coder","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kat-dev-32b":{"mode":"chat","base_model":"kat-dev-32b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/kat-dev-72b-exp":{"mode":"chat","base_model":"kat-dev-72b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-guard-2-8b":{"mode":"chat","base_model":"llama-guard-2-8b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-guard-3-1b":{"mode":"chat","base_model":"llama-guard-3-1b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-guard-3-8b":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-guard-3-8b","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/llama-v2-13b":{"mode":"chat","base_model":"llama-2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v2-13b-chat":{"mode":"chat","base_model":"llama-2-13b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v2-70b":{"mode":"chat","base_model":"llama-2-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v2-70b-chat":{"mode":"chat","base_model":"llama-2-70b-chat","max_input_tokens":2048,"max_output_tokens":2048,"provider":"fireworks_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v2-7b":{"mode":"chat","base_model":"llama-2-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v2-7b-chat":{"mode":"chat","base_model":"llama-2-7b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3-70b-instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3-70b-instruct-hf":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3-8b":{"mode":"chat","base_model":"llama-3-8b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3-8b-instruct-hf":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p1-405b-instruct-long":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-v3p1-405b-instruct-long","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/llama-v3p1-70b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-v3p1-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/llama-v3p1-70b-instruct-1b":{"mode":"chat","base_model":"llama-3.1-70b-instruct-1b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p1-nemotron-70b-instruct":{"mode":"chat","base_model":"llama-3.1-nemotron-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-1b":{"mode":"chat","base_model":"llama-3.2-1b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p2-3b":{"mode":"chat","base_model":"llama-3.2-3b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llama-v3p3-70b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/llama-v3p3-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/llamaguard-7b":{"mode":"chat","base_model":"llamaguard-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/llava-yi-34b":{"mode":"chat","base_model":"llava-yi-34b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/minimax-m1-80k":{"mode":"chat","base_model":"minimax-m1-80k","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/minimax-m2":{"mode":"chat","base_model":"minimax-m2","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/ministral-3-14b-instruct-2512":{"mode":"chat","base_model":"ministral-3-14b-instruct","max_input_tokens":256000,"max_output_tokens":256000,"provider":"fireworks_ai","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/ministral-3-3b-instruct-2512":{"mode":"chat","base_model":"ministral-3-3b-instruct","max_input_tokens":256000,"max_output_tokens":256000,"provider":"fireworks_ai","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/ministral-3-8b-instruct-2512":{"mode":"chat","base_model":"ministral-3-8b-instruct","max_input_tokens":256000,"max_output_tokens":256000,"provider":"fireworks_ai","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-7b":{"mode":"chat","base_model":"mistral-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-4k":{"mode":"chat","base_model":"mistral-7b-instruct-4k","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-v0p2":{"mode":"chat","base_model":"mistral-7b-instruct-v0.2","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-v3":{"mode":"chat","base_model":"mistral-7b-instruct-v3","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-7b-v0p2":{"mode":"chat","base_model":"mistral-7b-v0.2","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-large-3-fp8":{"mode":"chat","base_model":"mistral-large-3-fp8","max_input_tokens":256000,"max_output_tokens":256000,"provider":"fireworks_ai","max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-nemo-base-2407":{"mode":"chat","base_model":"mistral-nemo-base","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-nemo-instruct-2407":{"mode":"chat","base_model":"mistral-nemo-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mistral-small-24b-instruct-2501":{"mode":"chat","base_model":"mistral-small-24b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mixtral-8x22b":{"mode":"chat","base_model":"mixtral-8x22b","max_input_tokens":65536,"max_output_tokens":65536,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mixtral-8x22b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/mixtral-8x22b-instruct","max_input_tokens":64000,"max_output_tokens":64000,"provider":"litellm","max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/mixtral-8x7b":{"mode":"chat","base_model":"mixtral-8x7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mixtral-8x7b-instruct":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mixtral-8x7b-instruct-hf":{"mode":"chat","base_model":"mixtral-8x7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/mythomax-l2-13b":{"mode":"chat","base_model":"mythomax-l2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nemotron-nano-v2-12b-vl":{"mode":"chat","base_model":"nemotron-nano-v2-12b-vl","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-capybara-7b-v1p9":{"mode":"chat","base_model":"nous-capybara-7b-v1.9","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-hermes-2-mixtral-8x7b-dpo":{"mode":"chat","base_model":"nous-hermes-2-mixtral-8x7b-dpo","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-hermes-2-yi-34b":{"mode":"chat","base_model":"nous-hermes-2-yi-34b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-13b":{"mode":"chat","base_model":"nous-hermes-llama2-13b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-70b":{"mode":"chat","base_model":"nous-hermes-llama2-70b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-7b":{"mode":"chat","base_model":"nous-hermes-llama2-7b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nvidia-nemotron-nano-12b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-12b-v2","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/nvidia-nemotron-nano-9b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-9b-v2","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/openchat-3p5-0106-7b":{"mode":"chat","base_model":"openchat-3.5-0106-7b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/openhermes-2-mistral-7b":{"mode":"chat","base_model":"openhermes-2-mistral-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/openhermes-2p5-mistral-7b":{"mode":"chat","base_model":"openhermes-2.5-mistral-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/openorca-7b":{"mode":"chat","base_model":"openorca-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phi-2-3b":{"mode":"chat","base_model":"phi-2-3b","max_input_tokens":2048,"max_output_tokens":2048,"provider":"fireworks_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phi-3-mini-128k-instruct":{"mode":"chat","base_model":"phi-3-mini-128k-instruct","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phi-3-vision-128k-instruct":{"mode":"chat","base_model":"phi-3-vision-128k-instruct","max_input_tokens":32064,"max_output_tokens":32064,"provider":"fireworks_ai","max_tokens":32064,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32064},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-python-v1":{"mode":"chat","base_model":"phind-code-llama-34b-python-v1","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-v1":{"mode":"chat","base_model":"phind-code-llama-34b-v1","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-v2":{"mode":"chat","base_model":"phind-code-llama-34b-v2","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/playground-v2-1024px-aesthetic":{"mode":"image_generation","base_model":"playground-v2-1024px-aesthetic","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/playground-v2-5-1024px-aesthetic":{"mode":"image_generation","base_model":"playground-v2-5-1024px-aesthetic","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/pythia-12b":{"mode":"chat","base_model":"pythia-12b","max_input_tokens":2048,"max_output_tokens":2048,"provider":"fireworks_ai","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen-qwq-32b-preview":{"mode":"chat","base_model":"qwq-32b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen-v2p5-14b-instruct":{"mode":"chat","base_model":"qwen2.5-14b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen-v2p5-7b":{"mode":"chat","base_model":"qwen2.5-7b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen1p5-72b-chat":{"mode":"chat","base_model":"qwen1.5-72b-chat","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2-7b-instruct":{"mode":"chat","base_model":"qwen2-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2-vl-2b-instruct":{"mode":"chat","base_model":"qwen2-vl-2b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2-vl-72b-instruct":{"mode":"image_generation","base_model":"fireworks/accounts/fireworks/models/qwen2-vl-72b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"provider":"litellm","max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/qwen2-vl-7b-instruct":{"mode":"chat","base_model":"qwen2-vl-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-0p5b-instruct":{"mode":"chat","base_model":"qwen2.5-0.5b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-14b":{"mode":"chat","base_model":"qwen2.5-14b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-1p5b-instruct":{"mode":"chat","base_model":"qwen2.5-1.5b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-32b":{"mode":"chat","base_model":"qwen2.5-32b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-32b-instruct":{"mode":"chat","base_model":"qwen2.5-32b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-72b":{"mode":"chat","base_model":"qwen2.5-72b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-72b-instruct":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/qwen2p5-72b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"provider":"litellm","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/qwen2p5-7b-instruct":{"mode":"chat","base_model":"qwen2.5-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-0p5b":{"mode":"chat","base_model":"qwen2.5-coder-0.5b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-0p5b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-0.5b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-14b":{"mode":"chat","base_model":"qwen2.5-coder-14b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-14b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-14b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-1p5b":{"mode":"chat","base_model":"qwen2.5-coder-1.5b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-1p5b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-1.5b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b":{"mode":"chat","base_model":"qwen2.5-coder-32b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-128k":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct-128k","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-32k-rope":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct-32k-rope","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-64k":{"mode":"chat","base_model":"qwen2.5-coder-32b-instruct-64k","max_input_tokens":65536,"max_output_tokens":65536,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-3b":{"mode":"chat","base_model":"qwen2.5-coder-3b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-3b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-3b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-7b":{"mode":"chat","base_model":"qwen2.5-coder-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-coder-7b-instruct":{"mode":"chat","base_model":"qwen2.5-coder-7b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-math-72b-instruct":{"mode":"chat","base_model":"qwen2.5-math-72b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-vl-32b-instruct":{"mode":"image_generation","base_model":"fireworks/accounts/fireworks/models/qwen2p5-vl-32b-instruct","max_input_tokens":125000,"max_output_tokens":125000,"provider":"litellm","max_tokens":125000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":125000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/qwen2p5-vl-3b-instruct":{"mode":"chat","base_model":"qwen2.5-vl-3b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-vl-72b-instruct":{"mode":"chat","base_model":"qwen2.5-vl-72b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen2p5-vl-7b-instruct":{"mode":"chat","base_model":"qwen2.5-vl-7b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-0p6b":{"mode":"chat","base_model":"qwen3-0.6b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-14b":{"mode":"chat","base_model":"qwen3-14b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-1p7b":{"mode":"chat","base_model":"qwen3-1.7b","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft":{"mode":"chat","base_model":"qwen3-1.7b-fp8-draft","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft-131072":{"mode":"chat","base_model":"qwen3-1.7b-fp8-draft-131072","max_input_tokens":131072,"max_output_tokens":131072,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft-40960":{"mode":"chat","base_model":"qwen3-1.7b-fp8-draft-40960","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/qwen3-235b-a22b","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b-instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b-thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/qwen3-30b-a3b","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b-instruct-2507":{"mode":"chat","base_model":"qwen3-30b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b-thinking-2507":{"mode":"chat","base_model":"qwen3-30b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-32b":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":131072,"max_output_tokens":131072,"supports_reasoning":true,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-4b":{"mode":"chat","base_model":"qwen3-4b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-4b-instruct-2507":{"mode":"chat","base_model":"qwen3-4b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-8b":{"mode":"chat","base_model":"qwen3-8b","max_input_tokens":40960,"max_output_tokens":40960,"supports_reasoning":true,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-coder-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-coder-480b-instruct-bf16":{"mode":"chat","base_model":"qwen3-coder-480b-instruct-bf16","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-embedding-0p6b":{"mode":"embedding","base_model":"qwen3-embedding-0.6b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768},"fireworks_ai/accounts/fireworks/models/qwen3-embedding-4b":{"mode":"embedding","base_model":"qwen3-embedding-4b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960},"fireworks_ai/accounts/fireworks/models/":{"mode":"embedding","base_model":"","max_tokens":40960,"max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai"},"fireworks_ai/accounts/fireworks/models/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-reranker-0p6b":{"mode":"rerank","base_model":"qwen3-reranker-0.6b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-reranker-4b":{"mode":"rerank","base_model":"qwen3-reranker-4b","max_input_tokens":40960,"max_output_tokens":40960,"provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-reranker-8b":{"mode":"rerank","base_model":"qwen3-reranker-8b","max_input_tokens":40960,"max_output_tokens":40960,"source":"https://api.fireworks.ai/v1/serverless/models?format=nested","provider":"fireworks_ai","max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-30b-a3b-thinking":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"provider":"fireworks_ai","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-32b-instruct":{"mode":"chat","base_model":"qwen3-vl-32b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-vl-8b-instruct":{"mode":"chat","base_model":"qwen3-vl-8b-instruct","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwq-32b":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/qwq-32b","max_input_tokens":128000,"max_output_tokens":128000,"provider":"litellm","max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/rolm-ocr":{"mode":"chat","base_model":"rolm-ocr","max_input_tokens":128000,"max_output_tokens":128000,"provider":"fireworks_ai","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/snorkel-mistral-7b-pairrm-dpo":{"mode":"chat","base_model":"snorkel-mistral-7b-pairrm-dpo","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/stable-diffusion-xl-1024-v1-0":{"mode":"image_generation","base_model":"stable-diffusion-xl-1024-v1-0","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/stablecode-3b":{"mode":"chat","base_model":"stablecode-3b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/starcoder-16b":{"mode":"chat","base_model":"starcoder-16b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/starcoder-7b":{"mode":"chat","base_model":"starcoder-7b","max_input_tokens":8192,"max_output_tokens":8192,"provider":"fireworks_ai","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/starcoder2-15b":{"mode":"chat","base_model":"starcoder2-15b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/starcoder2-3b":{"mode":"chat","base_model":"starcoder2-3b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/starcoder2-7b":{"mode":"chat","base_model":"starcoder2-7b","max_input_tokens":16384,"max_output_tokens":16384,"provider":"fireworks_ai","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/toppy-m-7b":{"mode":"chat","base_model":"toppy-m-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/whisper-v3":{"mode":"audio_transcription","base_model":"whisper-v3","max_tokens":4096,"max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/whisper-v3-turbo":{"mode":"audio_transcription","base_model":"whisper-v3-turbo","max_tokens":4096,"max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/yi-34b":{"mode":"chat","base_model":"yi-34b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/yi-34b-200k-capybara":{"mode":"chat","base_model":"yi-34b-200k-capybara","max_input_tokens":200000,"max_output_tokens":200000,"provider":"fireworks_ai","max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/yi-34b-chat":{"mode":"chat","base_model":"yi-34b-chat","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/yi-6b":{"mode":"chat","base_model":"yi-6b","max_input_tokens":4096,"max_output_tokens":4096,"provider":"fireworks_ai","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/zephyr-7b-beta":{"mode":"chat","base_model":"zephyr-7b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3.2":{"mode":"chat","base_model":"deepseek","max_input_tokens":163840,"max_output_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/minimax/minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","max_input_tokens":204800,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.7":{"mode":"chat","base_model":"glm-4.7","max_input_tokens":204800,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/xiaomimimo/mimo-v2-flash":{"mode":"chat","base_model":"mimo-v2-flash","max_input_tokens":262144,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/autoglm-phone-9b-multilingual":{"mode":"chat","base_model":"autoglm-phone-9b-multilingual","max_input_tokens":65536,"max_output_tokens":65536,"supports_vision":true,"supports_system_messages":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/moonshotai/kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/minimax/minimax-m2":{"mode":"chat","base_model":"minimax-m2","max_input_tokens":204800,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/paddlepaddle/paddleocr-vl":{"mode":"chat","base_model":"paddleocr-vl","max_input_tokens":16384,"max_output_tokens":16384,"supports_vision":true,"supports_system_messages":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3.2-exp":{"mode":"chat","base_model":"deepseek-v3.2","max_input_tokens":163840,"max_output_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"supports_vision":true,"supports_system_messages":true,"supports_reasoning":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.6v":{"mode":"chat","base_model":"glm-4.6v","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.6":{"mode":"chat","base_model":"glm-4.6","max_input_tokens":204800,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/kwaipilot/kat-coder-pro":{"mode":"chat","base_model":"kat-coder-pro","max_input_tokens":256000,"max_output_tokens":128000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-ocr":{"mode":"chat","base_model":"deepseek-ocr","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3.1-terminus":{"mode":"chat","base_model":"deepseek-v3.1-terminus","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-max":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":262144,"max_output_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/skywork/r1v4-lite":{"mode":"chat","base_model":"r1v4-lite","max_input_tokens":262144,"max_output_tokens":65536,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3.1":{"mode":"chat","base_model":"deepseek","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/moonshotai/kimi-k2-0905":{"mode":"chat","base_model":"kimi-k2","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-coder-480b-a35b-instruct":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","max_input_tokens":262144,"max_output_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-coder-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","max_input_tokens":160000,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/moonshotai/kimi-k2-instruct":{"mode":"chat","base_model":"kimi-k2-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.5":{"mode":"chat","base_model":"glm-4.5","max_input_tokens":131072,"max_output_tokens":98304,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":98304,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":98304},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-235b-a22b-thinking-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/meta-llama/llama-3.1-8b-instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":16384,"max_output_tokens":16384,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/google/gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","max_input_tokens":131072,"max_output_tokens":8192,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.5v":{"mode":"chat","base_model":"glm-4.5v","max_input_tokens":65536,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":32768,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/qwen/qwen3-235b-a22b-instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":131072,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-distill-qwen-14b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-14b","max_input_tokens":32768,"max_output_tokens":16384,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/meta-llama/llama-3.3-70b-instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":120000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":120000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":120000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen-2.5-72b-instruct":{"mode":"chat","base_model":"qwen2.5-72b-instruct","max_input_tokens":32000,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/mistralai/mistral-nemo":{"mode":"chat","base_model":"mistral-nemo","max_input_tokens":60288,"max_output_tokens":16000,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/minimaxai/minimax-m1-80k":{"mode":"chat","base_model":"minimax-m1-80k","max_input_tokens":1000000,"max_output_tokens":40000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"novita","max_tokens":40000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":163840,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"deepseek-r1-distill-qwen-32b","max_input_tokens":64000,"max_output_tokens":32000,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/meta-llama/llama-3-8b-instruct":{"mode":"chat","base_model":"llama-3-8b-instruct","max_input_tokens":8192,"max_output_tokens":8192,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/microsoft/wizardlm-2-8x22b":{"mode":"chat","base_model":"wizardlm-2-8x22b","max_input_tokens":65535,"max_output_tokens":8000,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-0528-qwen3-8b":{"mode":"chat","base_model":"deepseek-r1-0528-qwen3-8b","max_input_tokens":128000,"max_output_tokens":32000,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/meta-llama/llama-3-70b-instruct":{"mode":"chat","base_model":"llama-3-70b-instruct","max_input_tokens":8192,"max_output_tokens":8000,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/qwen/qwen3-235b-a22b-fp8":{"mode":"chat","base_model":"qwen3-235b-a22b-fp8","max_input_tokens":40960,"max_output_tokens":20000,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"novita","max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":1048576,"max_output_tokens":8192,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/meta-llama/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"llama-4-scout-17b-16e-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_vision":true,"supports_system_messages":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/nousresearch/hermes-2-pro-llama-3-8b":{"mode":"chat","base_model":"hermes-2-pro-llama-3-8b","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen2.5-vl-72b-instruct":{"mode":"chat","base_model":"qwen2.5-vl-72b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"supports_vision":true,"supports_system_messages":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/sao10k/l3-70b-euryale-v2.1":{"mode":"chat","base_model":"l3-70b-euryale","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-21B-a3b-thinking":{"mode":"chat","base_model":"ernie-4.5-21b-a3b-thinking","max_input_tokens":131072,"max_output_tokens":65536,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/sao10k/l3-8b-lunaris":{"mode":"chat","base_model":"l3-8b-lunaris","max_input_tokens":8192,"max_output_tokens":8192,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baichuan/baichuan-m2-32b":{"mode":"chat","base_model":"baichuan-m2-32b","max_input_tokens":131072,"max_output_tokens":131072,"supports_system_messages":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-vl-424b-a47b":{"mode":"chat","base_model":"ernie-4.5-vl-424b-a47b","max_input_tokens":123000,"max_output_tokens":16000,"supports_vision":true,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-300b-a47b-paddle":{"mode":"chat","base_model":"ernie-4.5-300b-a47b-paddle","max_input_tokens":123000,"max_output_tokens":12000,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":12000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":12000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-prover-v2-671b":{"mode":"chat","base_model":"deepseek-prover-v2-671b","max_input_tokens":160000,"max_output_tokens":160000,"supports_system_messages":true,"provider":"novita","max_tokens":160000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":160000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-32b-fp8":{"mode":"chat","base_model":"qwen3-32b-fp8","max_input_tokens":40960,"max_output_tokens":20000,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-30b-a3b-fp8":{"mode":"chat","base_model":"qwen3-30b-a3b-fp8","max_input_tokens":40960,"max_output_tokens":20000,"supports_system_messages":true,"supports_reasoning":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"novita","max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/google/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","max_input_tokens":98304,"max_output_tokens":16384,"supports_vision":true,"supports_system_messages":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-v3-turbo":{"mode":"chat","base_model":"deepseek-v3-turbo","max_input_tokens":64000,"max_output_tokens":16000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/deepseek/deepseek-r1-turbo":{"mode":"chat","base_model":"deepseek-r1-turbo","max_input_tokens":64000,"max_output_tokens":16000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/Sao10K/L3-8B-Stheno-v3.2":{"mode":"chat","base_model":"l3-8b-stheno","max_input_tokens":8192,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/gryphe/mythomax-l2-13b":{"mode":"chat","base_model":"mythomax-l2-13b","max_input_tokens":4096,"max_output_tokens":3200,"supports_system_messages":true,"provider":"novita","max_tokens":3200,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":3200},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-vl-28b-a3b-thinking":{"mode":"chat","base_model":"ernie-4.5-vl-28b-a3b-thinking","max_input_tokens":131072,"max_output_tokens":65536,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-vl-8b-instruct":{"mode":"chat","base_model":"qwen3-vl-8b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/zai-org/glm-4.5-air":{"mode":"chat","base_model":"glm-4.5-air","max_input_tokens":131072,"max_output_tokens":98304,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"provider":"novita","max_tokens":98304,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":98304},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-vl-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-vl-30b-a3b-thinking":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-omni-30b-a3b-thinking":{"mode":"chat","base_model":"qwen3-omni-30b-a3b-thinking","max_input_tokens":65536,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_reasoning":true,"supports_audio_input":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-omni-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-omni-30b-a3b-instruct","max_input_tokens":65536,"max_output_tokens":16384,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_response_schema":true,"supports_audio_input":true,"supports_audio_output":true,"provider":"novita","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen-mt-plus":{"mode":"chat","base_model":"qwen-mt-plus","max_input_tokens":16384,"max_output_tokens":8192,"supports_system_messages":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-vl-28b-a3b":{"mode":"chat","base_model":"ernie-4.5-vl-28b-a3b","max_input_tokens":30000,"max_output_tokens":8000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baidu/ernie-4.5-21B-a3b":{"mode":"chat","base_model":"ernie-4.5-21b-a3b","max_input_tokens":120000,"max_output_tokens":8000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-8b-fp8":{"mode":"chat","base_model":"qwen3-8b-fp8","max_input_tokens":128000,"max_output_tokens":20000,"supports_system_messages":true,"supports_reasoning":true,"provider":"novita","max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-4b-fp8":{"mode":"chat","base_model":"qwen3-4b-fp8","max_input_tokens":128000,"max_output_tokens":20000,"supports_system_messages":true,"supports_reasoning":true,"supports_function_calling":true,"supports_tool_choice":true,"provider":"novita","max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen2.5-7b-instruct":{"mode":"chat","base_model":"qwen2.5-7b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"novita/meta-llama/llama-3.2-3b-instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","max_input_tokens":32768,"max_output_tokens":32000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/sao10k/l31-70b-euryale-v2.2":{"mode":"chat","base_model":"l31-70b-euryale","max_input_tokens":8192,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/qwen/qwen3-embedding-0.6b":{"mode":"embedding","base_model":"qwen3-embedding-0.6b","max_input_tokens":32768,"max_output_tokens":32768,"provider":"novita","max_tokens":32768},"novita/qwen/qwen3-embedding-8b":{"mode":"embedding","base_model":"qwen3-embedding-8b","max_input_tokens":32768,"max_output_tokens":4096,"provider":"novita","max_tokens":4096},"novita/baai/bge-m3":{"mode":"embedding","base_model":"bge-m3","max_input_tokens":8192,"max_output_tokens":96000,"provider":"novita","max_tokens":96000},"novita/qwen/qwen3-reranker-8b":{"mode":"rerank","base_model":"qwen3-reranker-8b","max_input_tokens":32768,"max_output_tokens":4096,"provider":"novita","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"novita/baai/bge-reranker-v2-m3":{"mode":"rerank","base_model":"bge-reranker-v2-m3","max_input_tokens":8000,"max_output_tokens":8000,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"llamagate/llama-3.1-8b":{"mode":"chat","base_model":"llama-3.1-8b","max_input_tokens":131072,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/llama-3.2-3b":{"mode":"chat","base_model":"llama-3.2-3b","max_input_tokens":131072,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/mistral-7b-v0.3":{"mode":"chat","base_model":"mistral-7b","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/qwen3-8b":{"mode":"chat","base_model":"qwen3-8b","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/dolphin3-8b":{"mode":"chat","base_model":"dolphin3-8b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/deepseek-r1-8b":{"mode":"chat","base_model":"deepseek-r1-8b","max_input_tokens":65536,"max_output_tokens":16384,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"llamagate","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/deepseek-r1-7b-qwen":{"mode":"chat","base_model":"deepseek-r1-7b-qwen","max_input_tokens":131072,"max_output_tokens":16384,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"llamagate","max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"llamagate/openthinker-7b":{"mode":"chat","base_model":"openthinker-7b","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/qwen2.5-coder-7b":{"mode":"chat","base_model":"qwen2.5-coder-7b","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/deepseek-coder-6.7b":{"mode":"chat","base_model":"deepseek-coder-6.7b","max_input_tokens":16384,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/codellama-7b":{"mode":"chat","base_model":"codellama-7b","max_input_tokens":16384,"max_output_tokens":4096,"supports_function_calling":true,"supports_response_schema":true,"provider":"llamagate","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/qwen3-vl-8b":{"mode":"chat","base_model":"qwen3-vl-8b","max_input_tokens":32768,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"llamagate/llava-7b":{"mode":"chat","base_model":"llava-7b","max_input_tokens":4096,"max_output_tokens":2048,"supports_response_schema":true,"supports_vision":true,"provider":"llamagate","max_tokens":2048,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/gemma3-4b":{"mode":"chat","base_model":"gemma-3-4b","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"llamagate","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"llamagate/nomic-embed-text":{"mode":"embedding","base_model":"nomic-embed-text","max_input_tokens":8192,"provider":"llamagate","max_tokens":8192},"llamagate/qwen3-embedding-8b":{"mode":"embedding","base_model":"qwen3-embedding-8b","max_input_tokens":40960,"provider":"llamagate","max_tokens":40960},"sarvam/sarvam-m":{"mode":"chat","base_model":"sarvam-m","max_input_tokens":8192,"max_output_tokens":32000,"supports_reasoning":true,"provider":"sarvam","is_deprecated":true,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tts-1-1106":{"mode":"audio_speech","base_model":"tts-1","supported_endpoints":["/v1/audio/speech"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"tts-1-hd-1106":{"mode":"audio_speech","base_model":"tts-1-hd","supported_endpoints":["/v1/audio/speech"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-tts-2025-03-20":{"mode":"audio_speech","base_model":"gpt-4o-mini-tts","deprecation_date":"2027-01-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/models/gpt-4o-mini-tts","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text","audio"],"supported_output_modalities":["audio"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-tts-2025-12-15":{"mode":"audio_speech","base_model":"gpt-4o-mini-tts","deprecation_date":"2027-01-06","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text","audio"],"supported_output_modalities":["audio"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-transcribe-2025-03-20":{"mode":"audio_transcription","base_model":"gpt-4o-mini-transcribe","deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-4o-mini-transcribe-2025-12-15","max_input_tokens":16000,"max_output_tokens":2000,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-4o-mini-transcribe-2025-12-15":{"mode":"audio_transcription","base_model":"gpt-4o-mini-transcribe","deprecation_date":"2027-02-26","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-live-transcribe","max_input_tokens":16000,"max_output_tokens":2000,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-5-search-api":{"mode":"chat","base_model":"gpt-5-search-api","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":false,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-5-search-api-2025-10-14":{"mode":"chat","base_model":"gpt-5-search-api","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":false,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_reasoning":false,"supports_service_tier":true,"supports_assistant_prefill":true},"gpt-realtime-mini-2025-10-06":{"mode":"chat","base_model":"gpt-realtime-mini","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gpt-realtime-mini-2025-12-15":{"mode":"chat","base_model":"gpt-realtime-mini","max_input_tokens":128000,"max_output_tokens":4096,"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"deprecation_date":"2027-01-20","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-realtime-2.1-mini","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sora-2":{"mode":"video_generation","base_model":"sora-2","deprecation_date":"2026-09-24","deprecation_source":"https://developers.openai.com/api/docs/deprecations","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"supported_resolutions":["720x1280","1280x720"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sora-2-pro":{"mode":"video_generation","base_model":"sora-2-pro","deprecation_date":"2026-09-24","deprecation_source":"https://developers.openai.com/api/docs/deprecations","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"supported_resolutions":["720x1280","1280x720"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"sora-2-pro-high-res":{"mode":"video_generation","base_model":"sora-2-pro","deprecation_date":"2026-09-24","deprecation_source":"https://developers.openai.com/api/docs/deprecations","source":"https://platform.openai.com/docs/api-reference/videos","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"openai","web_search_billing_unit":"per_request","is_deprecated":true,"supported_resolutions":["1024x1792","1792x1024"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"chatgpt-image-latest":{"mode":"image_generation","base_model":"chatgpt-image","deprecation_date":"2026-12-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-image-2.6-sunburst","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.0-flash-exp-image-generation":{"mode":"image_generation","base_model":"gemini-2.0-flash-exp-image-generation","max_images_per_prompt":3000,"max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"source":"https://ai.google.dev/pricing","supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_vision":true,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.0-flash-exp-image-generation":{"mode":"image_generation","base_model":"gemini-2.0-flash-exp-image-generation","max_images_per_prompt":3000,"max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"source":"https://ai.google.dev/pricing","supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_vision":true,"tpm":250000,"rpm":10,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.0-flash-lite-001":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-03-31","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":8192,"max_pdf_size_mb":50,"max_video_length":1,"max_videos_per_prompt":10,"rpm":4000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-2.0-flash-lite","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"is_deprecated":true},"gemini-2.5-flash-native-audio-latest":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-native-audio-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-native-audio-preview-12-2025":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":8192,"max_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_response_schema":false,"supports_vision":true,"supports_web_search":true,"supports_reasoning":true},"gemini/gemini-2.5-flash-native-audio-latest":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"tpm":250000,"rpm":10,"gemini_native_audio":true,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-native-audio-preview-09-2025":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","max_input_tokens":1048576,"max_output_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"tpm":250000,"rpm":10,"gemini_native_audio":true,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini/gemini-2.5-flash-native-audio-preview-12-2025":{"mode":"chat","base_model":"gemini-2.5-flash-native-audio","deprecation_date":"2027-03-31","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-live","max_input_tokens":1048576,"max_output_tokens":8192,"source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"tpm":250000,"rpm":10,"gemini_native_audio":true,"supports_function_calling":true,"supports_response_schema":false,"supports_vision":true,"supports_web_search":true,"supports_reasoning":true,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-2.5-flash-preview-tts":{"mode":"audio_speech","base_model":"gemini-2.5-flash-tts","source":"https://ai.google.dev/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_audio_input":false,"supports_function_calling":false,"supports_response_schema":false,"supports_web_search":false},"gemini-flash-latest":{"mode":"chat","base_model":"gemini-flash","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"provider":"gemini","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true,"supports_audio_input":true,"supports_native_streaming":true,"supports_video_input":true},"gemini-flash-lite-latest":{"mode":"chat","base_model":"gemini-flash-lite","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":15,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"provider":"gemini","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true,"supports_audio_input":true,"supports_native_streaming":true,"supports_video_input":true},"gemini-pro-latest":{"mode":"chat","base_model":"gemini-pro","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":2000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"provider":"gemini","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true,"supports_native_streaming":true,"supports_url_context":true},"gemini/gemini-pro-latest":{"mode":"chat","base_model":"gemini-pro","max_input_tokens":1048576,"max_output_tokens":65535,"rpm":2000,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"prompt_cache_min_tokens":4096,"supports_native_streaming":true,"supports_url_context":true,"web_search_billing_unit":"per_query","provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-exp-1206":{"mode":"chat","base_model":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_input_tokens":1048576,"max_output_tokens":65535,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"rpm":100000,"source":"https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"tpm":8000000,"provider":"gemini","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"cerebras/qwen-3-235b-a22b-instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_tokens":131072,"max_input_tokens":131072,"max_output_tokens":131072,"provider":"cerebras","supports_function_calling":true,"supports_vision":false,"supports_reasoning":false,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"fireworks_ai/accounts/fireworks/models/qwen3-embedding-8b":{"mode":"embedding","base_model":"qwen3-embedding-8b","max_tokens":32768,"max_input_tokens":32768,"max_output_tokens":0,"provider":"fireworks_ai"},"zero-one-ai/Yi-34B-Chat":{"mode":"chat","base_model":"zero-one-ai/Yi-34B-Chat","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","accesorKey":"type","default":{"type":"medium"},"options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"Austism/chronos-hermes-13b":{"mode":"chat","base_model":"Austism/chronos-hermes-13b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"DiscoResearch/DiscoLM-mixtral-8x7b-v2":{"mode":"chat","base_model":"DiscoResearch/DiscoLM-mixtral-8x7b-v2","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Gryphe/MythoMax-L2-13b":{"mode":"chat","base_model":"Gryphe/MythoMax-L2-13b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_function_calling":true},"lmsys/vicuna-13b-v1.5":{"mode":"chat","base_model":"lmsys/vicuna-13b-v1.5","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"lmsys/vicuna-7b-v1.5":{"mode":"chat","base_model":"lmsys/vicuna-7b-v1.5","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"lmsys/vicuna-13b-v1.5-16k":{"mode":"chat","base_model":"lmsys/vicuna-13b-v1.5-16k","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-13b-Instruct-hf":{"mode":"chat","base_model":"codellama/CodeLlama-13b-Instruct-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-34b-Instruct-hf":{"mode":"chat","base_model":"codellama/CodeLlama-34b-Instruct-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-70b-Instruct-hf":{"mode":"chat","base_model":"codellama/CodeLlama-70b-Instruct-hf","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-7b-Instruct-hf":{"mode":"chat","base_model":"codellama/CodeLlama-7b-Instruct-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/llama-2-13b-chat":{"mode":"chat","base_model":"togethercomputer/llama-2-13b-chat","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_tool_choice":true},"togethercomputer/llama-2-70b-chat":{"mode":"chat","base_model":"togethercomputer/llama-2-70b-chat","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/llama-2-7b-chat":{"mode":"chat","base_model":"togethercomputer/llama-2-7b-chat","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_tool_choice":true},"NousResearch/Nous-Capybara-7B-V1p9":{"mode":"chat","base_model":"NousResearch/Nous-Capybara-7B-V1p9","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-Llama2-70b":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-Llama2-70b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-llama-2-7b":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-llama-2-7b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-Llama2-13b":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-Llama2-13b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-2-Yi-34B":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-2-Yi-34B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openchat/openchat-3.5-1210":{"mode":"chat","base_model":"openchat/openchat-3.5-1210","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Open-Orca/Mistral-7B-OpenOrca":{"mode":"chat","base_model":"Open-Orca/Mistral-7B-OpenOrca","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Qwen-7B-Chat":{"mode":"chat","base_model":"togethercomputer/Qwen-7B-Chat","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"snorkelai/Snorkel-Mistral-PairRM-DPO":{"mode":"chat","base_model":"snorkelai/Snorkel-Mistral-PairRM-DPO","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/alpaca-7b":{"mode":"chat","base_model":"togethercomputer/alpaca-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/falcon-40b-instruct":{"mode":"chat","base_model":"togethercomputer/falcon-40b-instruct","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/falcon-7b-instruct":{"mode":"chat","base_model":"togethercomputer/falcon-7b-instruct","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/GPT-NeoXT-Chat-Base-20B":{"mode":"chat","base_model":"togethercomputer/GPT-NeoXT-Chat-Base-20B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Llama-2-7B-32K-Instruct":{"mode":"chat","base_model":"togethercomputer/Llama-2-7B-32K-Instruct","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Pythia-Chat-Base-7B-v0.16":{"mode":"chat","base_model":"togethercomputer/Pythia-Chat-Base-7B-v0.16","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-Chat-3B-v1":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-Chat-3B-v1","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-7B-Chat":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-7B-Chat","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/StripedHyena-Nous-7B":{"mode":"chat","base_model":"togethercomputer/StripedHyena-Nous-7B","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Undi95/ReMM-SLERP-L2-13B":{"mode":"chat","base_model":"Undi95/ReMM-SLERP-L2-13B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true},"Undi95/Toppy-M-7B":{"mode":"chat","base_model":"Undi95/Toppy-M-7B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"WizardLM/WizardLM-13B-V1.2":{"mode":"chat","base_model":"WizardLM/WizardLM-13B-V1.2","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"garage-bAInd/Platypus2-70B-instruct":{"mode":"chat","base_model":"garage-bAInd/Platypus2-70B-instruct","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/Mistral-7B-Instruct-v0.2":{"mode":"chat","base_model":"mistralai/Mistral-7B-Instruct-v0.2","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"teknium/OpenHermes-2-Mistral-7B":{"mode":"chat","base_model":"teknium/OpenHermes-2-Mistral-7B","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"teknium/OpenHermes-2p5-Mistral-7B":{"mode":"chat","base_model":"teknium/OpenHermes-2p5-Mistral-7B","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"upstage/SOLAR-10.7B-Instruct-v1.0":{"mode":"chat","base_model":"upstage/SOLAR-10.7B-Instruct-v1.0","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"zero-one-ai/Yi-34B":{"mode":"chat","base_model":"zero-one-ai/Yi-34B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"zero-one-ai/Yi-6B":{"mode":"chat","base_model":"zero-one-ai/Yi-6B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/llemma_7b":{"mode":"chat","base_model":"EleutherAI/llemma_7b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"huggyllama/llama-65b":{"mode":"chat","base_model":"huggyllama/llama-65b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/llama-2-13b":{"mode":"chat","base_model":"togethercomputer/llama-2-13b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"togethercomputer/llama-2-70b":{"mode":"chat","base_model":"togethercomputer/llama-2-70b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"togethercomputer/llama-2-7b":{"mode":"chat","base_model":"togethercomputer/llama-2-7b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"microsoft/phi-2":{"mode":"chat","base_model":"microsoft/phi-2","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Nexusflow/NexusRaven-V2-13B":{"mode":"chat","base_model":"Nexusflow/NexusRaven-V2-13B","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Qwen-7B":{"mode":"chat","base_model":"togethercomputer/Qwen-7B","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/falcon-40b":{"mode":"chat","base_model":"togethercomputer/falcon-40b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/falcon-7b":{"mode":"chat","base_model":"togethercomputer/falcon-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/GPT-JT-6B-v1":{"mode":"chat","base_model":"togethercomputer/GPT-JT-6B-v1","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/GPT-JT-Moderation-6B":{"mode":"chat","base_model":"togethercomputer/GPT-JT-Moderation-6B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/LLaMA-2-7B-32K":{"mode":"chat","base_model":"togethercomputer/LLaMA-2-7B-32K","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-Base-3B-v1":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-Base-3B-v1","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-7B-Base":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-7B-Base","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-Instruct-3B-v1":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-Instruct-3B-v1","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/RedPajama-INCITE-7B-Instruct":{"mode":"chat","base_model":"togethercomputer/RedPajama-INCITE-7B-Instruct","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/StripedHyena-Hessian-7B":{"mode":"chat","base_model":"togethercomputer/StripedHyena-Hessian-7B","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"WizardLM/WizardLM-70B-V1.0":{"mode":"chat","base_model":"WizardLM/WizardLM-70B-V1.0","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/Mistral-7B-v0.1":{"mode":"chat","base_model":"mistralai/Mistral-7B-v0.1","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/Mixtral-8x7B-v0.1":{"mode":"chat","base_model":"mistralai/Mixtral-8x7B-v0.1","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-13b-hf":{"mode":"chat","base_model":"codellama/CodeLlama-13b-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-34b-hf":{"mode":"chat","base_model":"codellama/CodeLlama-34b-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-70b-hf":{"mode":"chat","base_model":"codellama/CodeLlama-70b-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-7b-hf":{"mode":"chat","base_model":"codellama/CodeLlama-7b-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-13b-Python-hf":{"mode":"chat","base_model":"codellama/CodeLlama-13b-Python-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-34b-Python-hf":{"mode":"chat","base_model":"codellama/CodeLlama-34b-Python-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-70b-Python-hf":{"mode":"chat","base_model":"codellama/CodeLlama-70b-Python-hf","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"codellama/CodeLlama-7b-Python-hf":{"mode":"chat","base_model":"codellama/CodeLlama-7b-Python-hf","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NumbersStation/nsql-llama-2-7B":{"mode":"chat","base_model":"NumbersStation/nsql-llama-2-7B","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Phind/Phind-CodeLlama-34B-Python-v1":{"mode":"chat","base_model":"Phind/Phind-CodeLlama-34B-Python-v1","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Phind/Phind-CodeLlama-34B-v2":{"mode":"chat","base_model":"Phind/Phind-CodeLlama-34B-v2","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"WizardLM/WizardCoder-Python-34B-V1.0":{"mode":"chat","base_model":"WizardLM/WizardCoder-Python-34B-V1.0","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"WizardLM/WizardCoder-15B-V1.0":{"mode":"chat","base_model":"WizardLM/WizardCoder-15B-V1.0","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"prompthero/openjourney":{"mode":"image_generation","base_model":"prompthero/openjourney","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"runwayml/stable-diffusion-v1-5":{"mode":"image_generation","base_model":"runwayml/stable-diffusion-v1-5","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"SG161222/Realistic_Vision_V3.0_VAE":{"mode":"image_generation","base_model":"SG161222/Realistic_Vision_V3.0_VAE","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"stabilityai/stable-diffusion-2-1":{"mode":"image_generation","base_model":"stabilityai/stable-diffusion-2-1","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"stabilityai/stable-diffusion-xl-base-1.0":{"mode":"image_generation","base_model":"stabilityai/stable-diffusion-xl-base-1.0","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"wavymulder/Analog-Diffusion":{"mode":"image_generation","base_model":"wavymulder/Analog-Diffusion","provider":"together_ai","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Meta-Llama/Llama-Guard-7b":{"mode":"moderation","base_model":"Meta-Llama/Llama-Guard-7b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"databricks/dolly-v2-12b":{"mode":"chat","base_model":"databricks/dolly-v2-12b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"databricks/dolly-v2-3b":{"mode":"chat","base_model":"databricks/dolly-v2-3b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"databricks/dolly-v2-7b":{"mode":"chat","base_model":"databricks/dolly-v2-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"HuggingFaceH4/zephyr-7b-beta":{"mode":"chat","base_model":"HuggingFaceH4/zephyr-7b-beta","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"HuggingFaceH4/starchat-alpha":{"mode":"chat","base_model":"HuggingFaceH4/starchat-alpha","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5":{"mode":"chat","base_model":"OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"OpenAssistant/stablelm-7b-sft-v7-epoch-3":{"mode":"chat","base_model":"OpenAssistant/stablelm-7b-sft-v7-epoch-3","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Koala-13B":{"mode":"chat","base_model":"togethercomputer/Koala-13B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/Koala-7B":{"mode":"chat","base_model":"togethercomputer/Koala-7B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"lmsys/vicuna-13b-v1.3":{"mode":"chat","base_model":"lmsys/vicuna-13b-v1.3","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"lmsys/vicuna-7b-v1.3":{"mode":"chat","base_model":"lmsys/vicuna-7b-v1.3","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"lmsys/fastchat-t5-3b-v1.0":{"mode":"chat","base_model":"lmsys/fastchat-t5-3b-v1.0","provider":"together_ai","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/mpt-30b-chat":{"mode":"chat","base_model":"togethercomputer/mpt-30b-chat","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/mpt-7b-chat":{"mode":"chat","base_model":"togethercomputer/mpt-7b-chat","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/guanaco-13b":{"mode":"chat","base_model":"togethercomputer/guanaco-13b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/guanaco-33b":{"mode":"chat","base_model":"togethercomputer/guanaco-33b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/guanaco-65b":{"mode":"chat","base_model":"togethercomputer/guanaco-65b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"togethercomputer/guanaco-7b":{"mode":"chat","base_model":"togethercomputer/guanaco-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"defog/sqlcoder":{"mode":"chat","base_model":"defog/sqlcoder","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/gpt-j-6b":{"mode":"chat","base_model":"EleutherAI/gpt-j-6b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/gpt-neox-20b":{"mode":"chat","base_model":"EleutherAI/gpt-neox-20b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/pythia-12b-v0":{"mode":"chat","base_model":"EleutherAI/pythia-12b-v0","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/pythia-1b-v0":{"mode":"chat","base_model":"EleutherAI/pythia-1b-v0","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/pythia-2.8b-v0":{"mode":"chat","base_model":"EleutherAI/pythia-2.8b-v0","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"EleutherAI/pythia-6.9b":{"mode":"chat","base_model":"EleutherAI/pythia-6.9b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/flan-t5-xl":{"mode":"chat","base_model":"google/flan-t5-xl","provider":"together_ai","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/flan-t5-xxl":{"mode":"chat","base_model":"google/flan-t5-xxl","provider":"together_ai","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"huggyllama/llama-13b":{"mode":"chat","base_model":"huggyllama/llama-13b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"huggyllama/llama-30b":{"mode":"chat","base_model":"huggyllama/llama-30b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"huggyllama/llama-7b":{"mode":"chat","base_model":"huggyllama/llama-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mosaicml/mpt-7b":{"mode":"chat","base_model":"mosaicml/mpt-7b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mosaicml/mpt-7b-instruct":{"mode":"chat","base_model":"mosaicml/mpt-7b-instruct","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-13b":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-13b","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NumbersStation/nsql-6B":{"mode":"chat","base_model":"NumbersStation/nsql-6B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"stabilityai/stablelm-base-alpha-3b":{"mode":"chat","base_model":"stabilityai/stablelm-base-alpha-3b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"stabilityai/stablelm-base-alpha-7b":{"mode":"chat","base_model":"stabilityai/stablelm-base-alpha-7b","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"bigcode/starcoder":{"mode":"chat","base_model":"bigcode/starcoder","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Replit-Code-v1 (3B)":{"mode":"chat","base_model":"Replit-Code-v1 (3B)","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Salesforce/codegen2-16B":{"mode":"chat","base_model":"Salesforce/codegen2-16B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"Salesforce/codegen2-7B":{"mode":"chat","base_model":"Salesforce/codegen2-7B","provider":"together_ai","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"gpt-4-turbo-vision-128k":{"mode":"image_generation","base_model":"gpt-4-turbo-vision-128k","provider":"azure","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"DeepSeek-R1-0528":{"mode":"chat","base_model":"DeepSeek-R1-0528","provider":"azure","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_system_messages":true,"supports_tool_choice":false},"DeepSeek-V3-0324":{"mode":"chat","base_model":"DeepSeek-V3-0324","provider":"azure","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"publishers/anthropic/models/claude-3-5-haiku":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-3-5-haiku","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_pdf_input":true},"publishers/anthropic/models/claude-3-7-sonnet":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-3-7-sonnet","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true},"publishers/anthropic/models/claude-opus-4":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-opus-4","provider":"vertex_ai","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true},"publishers/anthropic/models/claude-sonnet-4":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-sonnet-4","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true},"publishers/anthropic/models/claude-opus-4-1":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-opus-4-1","provider":"vertex_ai","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"publishers/anthropic/models/claude-sonnet-4-5":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-sonnet-4-5","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true},"publishers/anthropic/models/claude-haiku-4-5":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-haiku-4-5","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"label":"High","value":"high"},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_native_structured_output":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true},"publishers/anthropic/models/claude-opus-4-5":{"mode":"image_generation","base_model":"publishers/anthropic/models/claude-opus-4-5","provider":"vertex_ai","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"anthropic_version","label":"Anthropic Version","helpText":"The version of the Anthropic API to use. default is vertex-2023-10-16","type":"text","default":"vertex-2023-10-16"}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_output_config":true,"supports_web_search":true},"publishers/google/models/gemini-2.5-pro":{"mode":"image_generation","base_model":"publishers/google/models/gemini-2.5-pro","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_url_context":true},"publishers/google/models/gemini-2.0-flash-001":{"mode":"chat","base_model":"publishers/google/models/gemini-2.0-flash-001","provider":"vertex_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"publishers/google/models/gemini-2.0-flash-lite-001":{"mode":"chat","base_model":"publishers/google/models/gemini-2.0-flash-lite-001","provider":"vertex_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"publishers/google/models/gemini-2.5-flash-lite":{"mode":"chat","base_model":"publishers/google/models/gemini-2.5-flash-lite","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_url_context":true,"supports_web_search":true,"supports_audio_input":true,"supports_image_size":false},"publishers/google/models/gemini-2.5-flash":{"mode":"chat","base_model":"publishers/google/models/gemini-2.5-flash","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_url_context":true,"supports_web_search":true,"supports_audio_input":true,"supports_image_size":false},"publishers/google/models/gemini-3-pro-preview":{"mode":"chat","base_model":"publishers/google/models/gemini-3-pro-preview","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_web_search":true},"publishers/google/models/gemini-3-flash-preview":{"mode":"chat","base_model":"publishers/google/models/gemini-3-flash-preview","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_url_context":true,"supports_web_search":true,"supports_native_streaming":true,"supports_audio_input":true,"supports_computer_use":true},"publishers/google/models/gemini-2.5-flash-image":{"mode":"chat","base_model":"publishers/google/models/gemini-2.5-flash-image","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_url_context":true,"supports_web_search":false,"supports_audio_input":false,"supports_image_size":false},"publishers/google/models/gemini-3-pro-image-preview":{"mode":"chat","base_model":"publishers/google/models/gemini-3-pro-image-preview","provider":"vertex_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_prompt_caching":true,"supports_system_messages":true,"supports_web_search":true,"supports_reasoning":false},"publishers/google/models/gemini-2.5-flash-lite-preview-09-2025":{"mode":"image_generation","base_model":"publishers/google/models/gemini-2.5-flash-lite-preview-09-2025","provider":"vertex_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"supports_image_size":false},"publishers/google/models/gemini-2.5-flash-preview-09-2025":{"mode":"image_generation","base_model":"publishers/google/models/gemini-2.5-flash-preview-09-2025","provider":"vertex_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_audio_output":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_url_context":true,"supports_vision":true,"supports_web_search":true,"supports_image_size":false},"publishers/google/models/gemini-2.5-flash-image-preview":{"mode":"image_generation","base_model":"publishers/google/models/gemini-2.5-flash-image-preview","provider":"vertex_ai","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openai/gpt-oss-120b-maas":{"mode":"chat","base_model":"openai/gpt-oss-120b-maas","provider":"vertex_ai","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_output_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"supported_endpoints":["/v1/responses"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_pdf_input":false},"openai/gpt-oss-20b-maas":{"mode":"chat","base_model":"openai/gpt-oss-20b-maas","provider":"vertex_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_output_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"supported_endpoints":["/v1/responses"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_pdf_input":false},"llama3-8b-8192":{"mode":"chat","base_model":"llama3-8b-8192","provider":"groq","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-70b-8192":{"mode":"chat","base_model":"llama3-70b-8192","provider":"groq","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-3.1-70b-versatile":{"mode":"chat","base_model":"llama-3.1-70b-versatile","provider":"groq","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-3.1-405b-reasoning":{"mode":"chat","base_model":"llama-3.1-405b-reasoning","provider":"groq","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"Qwen/Qwen1.5-72B-Chat":{"mode":"chat","base_model":"Qwen/Qwen1.5-72B-Chat","provider":"together_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"NousResearch/Nous-Hermes-2-Mistral-7B-DPO":{"mode":"chat","base_model":"NousResearch/Nous-Hermes-2-Mistral-7B-DPO","provider":"together_ai","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"deepseek-ai/deepseek-coder-33b-instruct":{"mode":"chat","base_model":"deepseek-ai/deepseek-coder-33b-instruct","provider":"together_ai","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/gemma-7b-it":{"mode":"chat","base_model":"google/gemma-7b-it","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo":{"mode":"chat","base_model":"meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/Llama-3-8b-chat-hf":{"mode":"chat","base_model":"meta-llama/Llama-3-8b-chat-hf","provider":"together_ai","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo":{"mode":"image_generation","base_model":"meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo","provider":"together_ai","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"command-light-nightly":{"mode":"chat","base_model":"command-light-nightly","provider":"cohere","max_input_tokens":4000,"max_output_tokens":4000,"max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek.r1-v1:0":{"mode":"chat","base_model":"deepseek.r1-v1:0","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-r1.html","max_input_tokens":32768,"max_output_tokens":32768,"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_reasoning":true,"supports_tool_choice":false,"provider":"bedrock","max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek-llm-r1-distill-qwen-7b":{"mode":"chat","base_model":"deepseek-llm-r1-distill-qwen-7b","provider":"bedrock","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-llm-r1-distill-qwen-32b":{"mode":"chat","base_model":"deepseek-llm-r1-distill-qwen-32b","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-llm-r1-distill-qwen-14b":{"mode":"chat","base_model":"deepseek-llm-r1-distill-qwen-14b","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-llm-r1-distill-llama-8b":{"mode":"chat","base_model":"deepseek-llm-r1-distill-llama-8b","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-llm-r1-distill-llama-70b":{"mode":"chat","base_model":"deepseek-llm-r1-distill-llama-70b","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-coder-7b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-coder-7b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-coder-32b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-coder-32b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-7b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-7b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-72b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-72b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-32b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-32b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"huggingface-llm-qwen2-5-14b-instruct":{"mode":"chat","base_model":"huggingface-llm-qwen2-5-14b-instruct","provider":"bedrock","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"ministral-8b-latest":{"mode":"chat","base_model":"ministral-8b-latest","provider":"mistral","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131000,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_response_schema":true,"supports_vision":true},"mistral-saba-latest":{"mode":"chat","base_model":"mistral-saba-latest","provider":"mistral","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32000,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"ministral-3b-latest":{"mode":"chat","base_model":"ministral-3b-latest","provider":"mistral","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131000,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_response_schema":true,"supports_vision":true},"accounts/yi-01-ai/models/yi-large":{"mode":"chat","base_model":"accounts/yi-01-ai/models/yi-large","provider":"fireworks_ai","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false},"accounts/sentientfoundation/models/dobby-unhinged-llama-3-3-70b-new":{"mode":"chat","base_model":"accounts/sentientfoundation/models/dobby-unhinged-llama-3-3-70b-new","provider":"fireworks_ai","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"accounts/fireworks/models/alpha":{"mode":"chat","base_model":"accounts/fireworks/models/alpha","provider":"fireworks_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"accounts/fireworks/models/moa":{"mode":"chat","base_model":"accounts/fireworks/models/moa","provider":"fireworks_ai","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/horizon-beta":{"mode":"image_generation","base_model":"openrouter/horizon-beta","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/codestral-2508":{"mode":"chat","base_model":"mistralai/codestral-2508","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true},"qwen/qwen3-30b-a3b-instruct-2507":{"mode":"chat","base_model":"qwen/qwen3-30b-a3b-instruct-2507","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true},"z-ai/glm-4.5-air:free":{"mode":"chat","base_model":"z-ai/glm-4.5-air:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"z-ai/glm-4-32b":{"mode":"chat","base_model":"z-ai/glm-4-32b","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-flash-lite":{"mode":"image_generation","base_model":"google/gemini-2.5-flash-lite","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"moonshotai/kimi-k2:free":{"mode":"chat","base_model":"moonshotai/kimi-k2:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moonshotai/kimi-k2":{"mode":"chat","base_model":"moonshotai/kimi-k2","provider":"openrouter","max_input_tokens":63000,"max_output_tokens":63000,"max_tokens":63000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":63000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"thudm/glm-4.1v-9b-thinking":{"mode":"image_generation","base_model":"thudm/glm-4.1v-9b-thinking","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/devstral-medium":{"mode":"chat","base_model":"mistralai/devstral-medium","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/devstral-small":{"mode":"chat","base_model":"mistralai/devstral-small","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"cognitivecomputations/dolphin-mistral-24b-venice-edition:free":{"mode":"chat","base_model":"cognitivecomputations/dolphin-mistral-24b-venice-edition:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/gemma-3n-e2b-it:free":{"mode":"chat","base_model":"google/gemma-3n-e2b-it:free","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"tencent/hunyuan-a13b-instruct:free":{"mode":"chat","base_model":"tencent/hunyuan-a13b-instruct:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"tencent/hunyuan-a13b-instruct":{"mode":"chat","base_model":"tencent/hunyuan-a13b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"tngtech/deepseek-r1t2-chimera:free":{"mode":"chat","base_model":"tngtech/deepseek-r1t2-chimera:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"baidu/ernie-4.5-300b-a47b":{"mode":"chat","base_model":"baidu/ernie-4.5-300b-a47b","provider":"openrouter","max_input_tokens":123000,"max_output_tokens":123000,"max_tokens":123000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":123000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"thedrummer/anubis-70b-v1.1":{"mode":"chat","base_model":"thedrummer/anubis-70b-v1.1","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"inception/mercury":{"mode":"chat","base_model":"inception/mercury","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/mistral-small-3.2-24b-instruct:free":{"mode":"image_generation","base_model":"mistralai/mistral-small-3.2-24b-instruct:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minimax/minimax-m1":{"mode":"chat","base_model":"minimax/minimax-m1","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-flash-lite-preview-06-17":{"mode":"image_generation","base_model":"google/gemini-2.5-flash-lite-preview-06-17","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"moonshotai/kimi-dev-72b:free":{"mode":"chat","base_model":"moonshotai/kimi-dev-72b:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/o3-pro":{"mode":"image_generation","base_model":"openai/o3-pro","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":200000,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true},"mistralai/magistral-small-2506":{"mode":"chat","base_model":"mistralai/magistral-small-2506","provider":"openrouter","max_input_tokens":40000,"max_output_tokens":40000,"max_tokens":40000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"mistralai/magistral-medium-2506":{"mode":"chat","base_model":"mistralai/magistral-medium-2506","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/magistral-medium-2506:thinking":{"mode":"chat","base_model":"mistralai/magistral-medium-2506:thinking","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-pro-preview":{"mode":"image_generation","base_model":"google/gemini-2.5-pro-preview","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"deepseek/deepseek-r1-distill-qwen-7b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-qwen-7b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-0528-qwen3-8b:free":{"mode":"chat","base_model":"deepseek/deepseek-r1-0528-qwen3-8b:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-0528-qwen3-8b":{"mode":"chat","base_model":"deepseek/deepseek-r1-0528-qwen3-8b","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true},"deepseek/deepseek-r1-0528:free":{"mode":"chat","base_model":"deepseek/deepseek-r1-0528:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"sarvamai/sarvam-m:free":{"mode":"chat","base_model":"sarvamai/sarvam-m:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"thedrummer/valkyrie-49b-v1":{"mode":"chat","base_model":"thedrummer/valkyrie-49b-v1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/devstral-small-2505:free":{"mode":"chat","base_model":"mistralai/devstral-small-2505:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/devstral-small-2505":{"mode":"chat","base_model":"mistralai/devstral-small-2505","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemma-3n-e4b-it:free":{"mode":"chat","base_model":"google/gemma-3n-e4b-it:free","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/gemma-3n-e4b-it":{"mode":"chat","base_model":"google/gemma-3n-e4b-it","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/codex-mini":{"mode":"image_generation","base_model":"openai/codex-mini","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true},"nousresearch/deephermes-3-mistral-24b-preview":{"mode":"chat","base_model":"nousresearch/deephermes-3-mistral-24b-preview","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/mistral-medium-3":{"mode":"image_generation","base_model":"mistralai/mistral-medium-3","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-pro-preview-05-06":{"mode":"image_generation","base_model":"google/gemini-2.5-pro-preview-05-06","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"arcee-ai/spotlight":{"mode":"image_generation","base_model":"arcee-ai/spotlight","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"arcee-ai/maestro-reasoning":{"mode":"chat","base_model":"arcee-ai/maestro-reasoning","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"arcee-ai/virtuoso-large":{"mode":"chat","base_model":"arcee-ai/virtuoso-large","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"arcee-ai/coder-large":{"mode":"chat","base_model":"arcee-ai/coder-large","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"microsoft/phi-4-reasoning-plus":{"mode":"chat","base_model":"microsoft/phi-4-reasoning-plus","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":false},"inception/mercury-coder":{"mode":"chat","base_model":"inception/mercury-coder","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen/qwen3-4b:free":{"mode":"chat","base_model":"qwen/qwen3-4b:free","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"opengvlab/internvl3-14b":{"mode":"image_generation","base_model":"opengvlab/internvl3-14b","provider":"openrouter","max_input_tokens":12288,"max_output_tokens":12288,"max_tokens":12288,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":12288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"deepseek/deepseek-prover-v2":{"mode":"chat","base_model":"deepseek/deepseek-prover-v2","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-guard-4-12b":{"mode":"image_generation","base_model":"meta-llama/llama-guard-4-12b","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_vision":true},"qwen/qwen3-30b-a3b:free":{"mode":"chat","base_model":"qwen/qwen3-30b-a3b:free","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwen3-30b-a3b":{"mode":"chat","base_model":"qwen/qwen3-30b-a3b","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"qwen/qwen3-8b:free":{"mode":"chat","base_model":"qwen/qwen3-8b:free","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwen3-8b":{"mode":"chat","base_model":"qwen/qwen3-8b","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_response_schema":true},"qwen/qwen3-14b:free":{"mode":"chat","base_model":"qwen/qwen3-14b:free","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwen3-14b":{"mode":"chat","base_model":"qwen/qwen3-14b","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"qwen/qwen3-32b":{"mode":"chat","base_model":"qwen/qwen3-32b","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"qwen/qwen3-235b-a22b:free":{"mode":"chat","base_model":"qwen/qwen3-235b-a22b:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwen3-235b-a22b":{"mode":"chat","base_model":"qwen/qwen3-235b-a22b","provider":"openrouter","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"tngtech/deepseek-r1t-chimera:free":{"mode":"chat","base_model":"tngtech/deepseek-r1t-chimera:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"tngtech/deepseek-r1t-chimera":{"mode":"chat","base_model":"tngtech/deepseek-r1t-chimera","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"microsoft/mai-ds-r1:free":{"mode":"chat","base_model":"microsoft/mai-ds-r1:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"microsoft/mai-ds-r1":{"mode":"chat","base_model":"microsoft/mai-ds-r1","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"thudm/glm-z1-32b:free":{"mode":"chat","base_model":"thudm/glm-z1-32b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"thudm/glm-4-32b":{"mode":"chat","base_model":"thudm/glm-4-32b","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/o4-mini-high":{"mode":"image_generation","base_model":"openai/o4-mini-high","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openai/o3":{"mode":"image_generation","base_model":"openai/o3","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":200000,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/o4-mini":{"mode":"image_generation","base_model":"openai/o4-mini","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":200000,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"shisa-ai/shisa-v2-llama3.3-70b:free":{"mode":"chat","base_model":"shisa-ai/shisa-v2-llama3.3-70b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"shisa-ai/shisa-v2-llama3.3-70b":{"mode":"chat","base_model":"shisa-ai/shisa-v2-llama3.3-70b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"alfredpros/codellama-7b-instruct-solidity":{"mode":"chat","base_model":"alfredpros/codellama-7b-instruct-solidity","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"arliai/qwq-32b-arliai-rpr-v1:free":{"mode":"chat","base_model":"arliai/qwq-32b-arliai-rpr-v1:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"arliai/qwq-32b-arliai-rpr-v1":{"mode":"chat","base_model":"arliai/qwq-32b-arliai-rpr-v1","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"agentica-org/deepcoder-14b-preview:free":{"mode":"chat","base_model":"agentica-org/deepcoder-14b-preview:free","provider":"openrouter","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"agentica-org/deepcoder-14b-preview":{"mode":"chat","base_model":"agentica-org/deepcoder-14b-preview","provider":"openrouter","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"moonshotai/kimi-vl-a3b-thinking:free":{"mode":"image_generation","base_model":"moonshotai/kimi-vl-a3b-thinking:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"moonshotai/kimi-vl-a3b-thinking":{"mode":"image_generation","base_model":"moonshotai/kimi-vl-a3b-thinking","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"nvidia/llama-3.3-nemotron-super-49b-v1":{"mode":"chat","base_model":"nvidia/llama-3.3-nemotron-super-49b-v1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"nvidia/llama-3.1-nemotron-ultra-253b-v1:free":{"mode":"chat","base_model":"nvidia/llama-3.1-nemotron-ultra-253b-v1:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"nvidia/llama-3.1-nemotron-ultra-253b-v1":{"mode":"chat","base_model":"nvidia/llama-3.1-nemotron-ultra-253b-v1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-4-maverick":{"mode":"image_generation","base_model":"meta-llama/llama-4-maverick","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"meta-llama/llama-4-scout":{"mode":"image_generation","base_model":"meta-llama/llama-4-scout","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true},"deepseek/deepseek-v3-base":{"mode":"chat","base_model":"deepseek/deepseek-v3-base","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"scb10x/llama3.1-typhoon2-70b-instruct":{"mode":"chat","base_model":"scb10x/llama3.1-typhoon2-70b-instruct","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/gemini-2.5-pro-exp-03-25":{"mode":"image_generation","base_model":"google/gemini-2.5-pro-exp-03-25","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwen2.5-vl-32b-instruct:free":{"mode":"image_generation","base_model":"qwen/qwen2.5-vl-32b-instruct:free","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen2.5-vl-32b-instruct":{"mode":"image_generation","base_model":"qwen/qwen2.5-vl-32b-instruct","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"deepseek/deepseek-chat-v3-0324:free":{"mode":"chat","base_model":"deepseek/deepseek-chat-v3-0324:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"featherless/qwerky-72b:free":{"mode":"chat","base_model":"featherless/qwerky-72b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/o1-pro":{"mode":"image_generation","base_model":"openai/o1-pro","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":200000,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true},"mistralai/mistral-small-3.1-24b-instruct:free":{"mode":"image_generation","base_model":"mistralai/mistral-small-3.1-24b-instruct:free","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemma-3-4b-it:free":{"mode":"image_generation","base_model":"google/gemma-3-4b-it:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"ai21/jamba-1.6-large":{"mode":"chat","base_model":"ai21/jamba-1.6-large","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"ai21/jamba-1.6-mini":{"mode":"chat","base_model":"ai21/jamba-1.6-mini","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"google/gemma-3-12b-it:free":{"mode":"image_generation","base_model":"google/gemma-3-12b-it:free","provider":"openrouter","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cohere/command-a":{"mode":"chat","base_model":"cohere/command-a","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"openai/gpt-4o-mini-search-preview":{"mode":"chat","base_model":"openai/gpt-4o-mini-search-preview","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"openai/gpt-4o-search-preview":{"mode":"chat","base_model":"openai/gpt-4o-search-preview","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"rekaai/reka-flash-3:free":{"mode":"chat","base_model":"rekaai/reka-flash-3:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"google/gemma-3-27b-it:free":{"mode":"image_generation","base_model":"google/gemma-3-27b-it:free","provider":"openrouter","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"thedrummer/anubis-pro-105b-v1":{"mode":"chat","base_model":"thedrummer/anubis-pro-105b-v1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"thedrummer/skyfall-36b-v2":{"mode":"chat","base_model":"thedrummer/skyfall-36b-v2","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_prompt_caching":true},"microsoft/phi-4-multimodal-instruct":{"mode":"image_generation","base_model":"microsoft/phi-4-multimodal-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwq-32b:free":{"mode":"chat","base_model":"qwen/qwq-32b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"qwen/qwq-32b":{"mode":"chat","base_model":"qwen/qwq-32b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"nousresearch/deephermes-3-llama-3-8b-preview:free":{"mode":"chat","base_model":"nousresearch/deephermes-3-llama-3-8b-preview:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash-lite-001":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-lite-001","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3.7-sonnet:thinking":{"mode":"image_generation","base_model":"anthropic/claude-3.7-sonnet:thinking","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3.7-sonnet:beta":{"mode":"image_generation","base_model":"anthropic/claude-3.7-sonnet:beta","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"perplexity/r1-1776":{"mode":"chat","base_model":"perplexity/r1-1776","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/mistral-saba":{"mode":"chat","base_model":"mistralai/mistral-saba","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true},"cognitivecomputations/dolphin3.0-r1-mistral-24b:free":{"mode":"chat","base_model":"cognitivecomputations/dolphin3.0-r1-mistral-24b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"cognitivecomputations/dolphin3.0-r1-mistral-24b":{"mode":"chat","base_model":"cognitivecomputations/dolphin3.0-r1-mistral-24b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"cognitivecomputations/dolphin3.0-mistral-24b:free":{"mode":"chat","base_model":"cognitivecomputations/dolphin3.0-mistral-24b:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cognitivecomputations/dolphin3.0-mistral-24b":{"mode":"chat","base_model":"cognitivecomputations/dolphin3.0-mistral-24b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-guard-3-8b":{"mode":"chat","base_model":"meta-llama/llama-guard-3-8b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false},"deepseek/deepseek-r1-distill-llama-8b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-llama-8b","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"aion-labs/aion-1.0":{"mode":"chat","base_model":"aion-labs/aion-1.0","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"aion-labs/aion-1.0-mini":{"mode":"chat","base_model":"aion-labs/aion-1.0-mini","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"aion-labs/aion-rp-llama-3.1-8b":{"mode":"chat","base_model":"aion-labs/aion-rp-llama-3.1-8b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen-vl-max":{"mode":"image_generation","base_model":"qwen/qwen-vl-max","provider":"openrouter","max_input_tokens":7500,"max_output_tokens":7500,"max_tokens":7500,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":7500},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen-turbo":{"mode":"chat","base_model":"qwen/qwen-turbo","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_reasoning":true},"qwen/qwen2.5-vl-72b-instruct:free":{"mode":"image_generation","base_model":"qwen/qwen2.5-vl-72b-instruct:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen2.5-vl-72b-instruct":{"mode":"image_generation","base_model":"qwen/qwen2.5-vl-72b-instruct","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"qwen/qwen-plus":{"mode":"chat","base_model":"qwen/qwen-plus","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_prompt_caching":true},"qwen/qwen-max":{"mode":"chat","base_model":"qwen/qwen-max","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-distill-qwen-1.5b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-qwen-1.5b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"mistralai/mistral-small-24b-instruct-2501:free":{"mode":"chat","base_model":"mistralai/mistral-small-24b-instruct-2501:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/mistral-small-24b-instruct-2501":{"mode":"chat","base_model":"mistralai/mistral-small-24b-instruct-2501","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"deepseek/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-qwen-32b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"deepseek/deepseek-r1-distill-qwen-14b:free":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-qwen-14b:free","provider":"openrouter","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-distill-qwen-14b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-qwen-14b","provider":"openrouter","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"liquid/lfm-7b":{"mode":"chat","base_model":"liquid/lfm-7b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"liquid/lfm-3b":{"mode":"chat","base_model":"liquid/lfm-3b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-distill-llama-70b:free":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-llama-70b:free","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"deepseek/deepseek-r1-distill-llama-70b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"deepseek/deepseek-r1:free":{"mode":"chat","base_model":"deepseek/deepseek-r1:free","provider":"openrouter","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"minimax/minimax-01":{"mode":"image_generation","base_model":"minimax/minimax-01","provider":"openrouter","max_input_tokens":1000192,"max_output_tokens":1000192,"max_tokens":1000192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/codestral-2501":{"mode":"chat","base_model":"mistralai/codestral-2501","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"microsoft/phi-4":{"mode":"chat","base_model":"microsoft/phi-4","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_function_calling":true},"sao10k/l3.3-euryale-70b":{"mode":"chat","base_model":"sao10k/l3.3-euryale-70b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cohere/command-r7b-12-2024":{"mode":"chat","base_model":"cohere/command-r7b-12-2024","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"google/gemini-2.0-flash-exp:free":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-exp:free","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.3-70b-instruct:free":{"mode":"chat","base_model":"meta-llama/llama-3.3-70b-instruct:free","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"amazon/nova-lite-v1":{"mode":"image_generation","base_model":"amazon/nova-lite-v1","provider":"openrouter","max_input_tokens":300000,"max_output_tokens":300000,"max_tokens":300000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":300000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"amazon/nova-micro-v1":{"mode":"chat","base_model":"amazon/nova-micro-v1","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"amazon/nova-pro-v1":{"mode":"image_generation","base_model":"amazon/nova-pro-v1","provider":"openrouter","max_input_tokens":300000,"max_output_tokens":300000,"max_tokens":300000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":300000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"qwen/qwq-32b-preview":{"mode":"chat","base_model":"qwen/qwq-32b-preview","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/gpt-4o-2024-11-20":{"mode":"image_generation","base_model":"openai/gpt-4o-2024-11-20","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"mistralai/mistral-large-2411":{"mode":"chat","base_model":"mistralai/mistral-large-2411","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"mistralai/mistral-large-2407":{"mode":"chat","base_model":"mistralai/mistral-large-2407","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true},"mistralai/pixtral-large-2411":{"mode":"image_generation","base_model":"mistralai/pixtral-large-2411","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"infermatic/mn-inferor-12b":{"mode":"chat","base_model":"infermatic/mn-inferor-12b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen-2.5-coder-32b-instruct:free":{"mode":"chat","base_model":"qwen/qwen-2.5-coder-32b-instruct:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"raifle/sorcererlm-8x22b":{"mode":"chat","base_model":"raifle/sorcererlm-8x22b","provider":"openrouter","max_input_tokens":16000,"max_output_tokens":16000,"max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"thedrummer/unslopnemo-12b":{"mode":"chat","base_model":"thedrummer/unslopnemo-12b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"anthracite-org/magnum-v4-72b":{"mode":"chat","base_model":"anthracite-org/magnum-v4-72b","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_prompt_caching":true},"mistralai/ministral-8b":{"mode":"chat","base_model":"mistralai/ministral-8b","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"mistralai/ministral-3b":{"mode":"chat","base_model":"mistralai/ministral-3b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true},"qwen/qwen-2.5-7b-instruct":{"mode":"chat","base_model":"qwen/qwen-2.5-7b-instruct","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"nvidia/llama-3.1-nemotron-70b-instruct":{"mode":"chat","base_model":"nvidia/llama-3.1-nemotron-70b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"inflection/inflection-3-productivity":{"mode":"chat","base_model":"inflection/inflection-3-productivity","provider":"openrouter","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"inflection/inflection-3-pi":{"mode":"chat","base_model":"inflection/inflection-3-pi","provider":"openrouter","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"thedrummer/rocinante-12b":{"mode":"chat","base_model":"thedrummer/rocinante-12b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"liquid/lfm-40b":{"mode":"chat","base_model":"liquid/lfm-40b","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"anthracite-org/magnum-v2-72b":{"mode":"chat","base_model":"anthracite-org/magnum-v2-72b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.2-3b-instruct:free":{"mode":"chat","base_model":"meta-llama/llama-3.2-3b-instruct:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.2-3b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3.2-3b-instruct","provider":"openrouter","max_input_tokens":20000,"max_output_tokens":20000,"max_tokens":20000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_system_messages":true},"meta-llama/llama-3.2-90b-vision-instruct":{"mode":"image_generation","base_model":"meta-llama/llama-3.2-90b-vision-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.2-1b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3.2-1b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":false},"meta-llama/llama-3.2-11b-vision-instruct:free":{"mode":"image_generation","base_model":"meta-llama/llama-3.2-11b-vision-instruct:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.2-11b-vision-instruct":{"mode":"image_generation","base_model":"meta-llama/llama-3.2-11b-vision-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":false,"supports_reasoning":false},"qwen/qwen-2.5-72b-instruct:free":{"mode":"chat","base_model":"qwen/qwen-2.5-72b-instruct:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen-2.5-72b-instruct":{"mode":"chat","base_model":"qwen/qwen-2.5-72b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_system_messages":true,"supports_response_schema":true},"neversleep/llama-3.1-lumimaid-8b":{"mode":"chat","base_model":"neversleep/llama-3.1-lumimaid-8b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/pixtral-12b":{"mode":"image_generation","base_model":"mistralai/pixtral-12b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"cohere/command-r-08-2024":{"mode":"chat","base_model":"cohere/command-r-08-2024","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"cohere/command-r-plus-08-2024":{"mode":"chat","base_model":"cohere/command-r-plus-08-2024","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"qwen/qwen-2.5-vl-7b-instruct":{"mode":"image_generation","base_model":"qwen/qwen-2.5-vl-7b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"sao10k/l3.1-euryale-70b":{"mode":"chat","base_model":"sao10k/l3.1-euryale-70b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true},"microsoft/phi-3.5-mini-128k-instruct":{"mode":"chat","base_model":"microsoft/phi-3.5-mini-128k-instruct","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"nousresearch/hermes-3-llama-3.1-70b":{"mode":"chat","base_model":"nousresearch/hermes-3-llama-3.1-70b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nousresearch/hermes-3-llama-3.1-405b":{"mode":"chat","base_model":"nousresearch/hermes-3-llama-3.1-405b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/chatgpt-4o-latest":{"mode":"image_generation","base_model":"openai/chatgpt-4o-latest","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"sao10k/l3-lunaris-8b":{"mode":"chat","base_model":"sao10k/l3-lunaris-8b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true},"openai/gpt-4o-2024-08-06":{"mode":"image_generation","base_model":"openai/gpt-4o-2024-08-06","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"meta-llama/llama-3.1-405b":{"mode":"chat","base_model":"meta-llama/llama-3.1-405b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.1-70b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3.1-70b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.1-405b-instruct:free":{"mode":"chat","base_model":"meta-llama/llama-3.1-405b-instruct:free","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.1-405b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3.1-405b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"meta-llama/llama-3.1-8b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3.1-8b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"mistralai/mistral-nemo:free":{"mode":"chat","base_model":"mistralai/mistral-nemo:free","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/mistral-nemo":{"mode":"chat","base_model":"mistralai/mistral-nemo","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"openai/gpt-4o-mini-2024-07-18":{"mode":"image_generation","base_model":"openai/gpt-4o-mini-2024-07-18","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/gpt-4o-mini":{"mode":"image_generation","base_model":"openai/gpt-4o-mini","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"google/gemma-2-27b-it":{"mode":"chat","base_model":"google/gemma-2-27b-it","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true},"google/gemma-2-9b-it:free":{"mode":"chat","base_model":"google/gemma-2-9b-it:free","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"google/gemma-2-9b-it":{"mode":"chat","base_model":"google/gemma-2-9b-it","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"sao10k/l3-euryale-70b":{"mode":"chat","base_model":"sao10k/l3-euryale-70b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cognitivecomputations/dolphin-mixtral-8x22b":{"mode":"chat","base_model":"cognitivecomputations/dolphin-mixtral-8x22b","provider":"openrouter","max_input_tokens":16000,"max_output_tokens":16000,"max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"qwen/qwen-2-72b-instruct":{"mode":"chat","base_model":"qwen/qwen-2-72b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"mistralai/mistral-7b-instruct-v0.3":{"mode":"chat","base_model":"mistralai/mistral-7b-instruct-v0.3","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nousresearch/hermes-2-pro-llama-3-8b":{"mode":"chat","base_model":"nousresearch/hermes-2-pro-llama-3-8b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"mistralai/mistral-7b-instruct:free":{"mode":"chat","base_model":"mistralai/mistral-7b-instruct:free","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"microsoft/phi-3-mini-128k-instruct":{"mode":"chat","base_model":"microsoft/phi-3-mini-128k-instruct","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"microsoft/phi-3-medium-128k-instruct":{"mode":"chat","base_model":"microsoft/phi-3-medium-128k-instruct","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"neversleep/llama-3-lumimaid-70b":{"mode":"chat","base_model":"neversleep/llama-3-lumimaid-70b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-guard-2-8b":{"mode":"chat","base_model":"meta-llama/llama-guard-2-8b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/gpt-4o:extended":{"mode":"image_generation","base_model":"openai/gpt-4o:extended","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sao10k/fimbulvetr-11b-v2":{"mode":"chat","base_model":"sao10k/fimbulvetr-11b-v2","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"meta-llama/llama-3-8b-instruct":{"mode":"chat","base_model":"meta-llama/llama-3-8b-instruct","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"microsoft/wizardlm-2-8x22b":{"mode":"chat","base_model":"microsoft/wizardlm-2-8x22b","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"openai/gpt-4-turbo":{"mode":"image_generation","base_model":"openai/gpt-4-turbo","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"cohere/command-r-plus":{"mode":"chat","base_model":"cohere/command-r-plus","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"cohere/command-r-plus-04-2024":{"mode":"chat","base_model":"cohere/command-r-plus-04-2024","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sophosympatheia/midnight-rose-70b":{"mode":"chat","base_model":"sophosympatheia/midnight-rose-70b","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cohere/command":{"mode":"chat","base_model":"cohere/command","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cohere/command-r":{"mode":"chat","base_model":"cohere/command-r","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-haiku:beta":{"mode":"image_generation","base_model":"anthropic/claude-3-haiku:beta","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-opus:beta":{"mode":"image_generation","base_model":"anthropic/claude-3-opus:beta","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-opus":{"mode":"image_generation","base_model":"anthropic/claude-3-opus","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"cohere/command-r-03-2024":{"mode":"chat","base_model":"cohere/command-r-03-2024","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-3.5-turbo-0613":{"mode":"chat","base_model":"openai/gpt-3.5-turbo-0613","provider":"openrouter","max_input_tokens":4095,"max_output_tokens":4095,"max_tokens":4095,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4095},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"openai/gpt-4-turbo-preview":{"mode":"chat","base_model":"openai/gpt-4-turbo-preview","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistralai/mistral-small":{"mode":"chat","base_model":"mistralai/mistral-small","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_reasoning":true,"supports_response_schema":true},"mistralai/mistral-tiny":{"mode":"chat","base_model":"mistralai/mistral-tiny","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_response_schema":true},"mistralai/mixtral-8x7b-instruct":{"mode":"chat","base_model":"mistralai/mixtral-8x7b-instruct","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neversleep/noromaid-20b":{"mode":"chat","base_model":"neversleep/noromaid-20b","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"alpindale/goliath-120b":{"mode":"chat","base_model":"alpindale/goliath-120b","provider":"openrouter","max_input_tokens":6144,"max_output_tokens":6144,"max_tokens":6144,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":6144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/auto":{"mode":"chat","base_model":"openrouter/auto","provider":"openrouter","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true},"openai/gpt-4-1106-preview":{"mode":"chat","base_model":"openai/gpt-4-1106-preview","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true},"mistralai/mistral-7b-instruct-v0.1":{"mode":"chat","base_model":"mistralai/mistral-7b-instruct-v0.1","provider":"openrouter","max_input_tokens":2824,"max_output_tokens":2824,"max_tokens":2824,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2824},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-3.5-turbo-instruct":{"mode":"chat","base_model":"openai/gpt-3.5-turbo-instruct","provider":"openrouter","max_input_tokens":4095,"max_output_tokens":4095,"max_tokens":4095,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4095},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"pygmalionai/mythalion-13b":{"mode":"chat","base_model":"pygmalionai/mythalion-13b","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openai/gpt-4-0314":{"mode":"chat","base_model":"openai/gpt-4-0314","provider":"openrouter","max_input_tokens":8191,"max_output_tokens":8191,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen-3-235b-a22b-thinking-2507":{"mode":"chat","base_model":"qwen-3-235b-a22b-thinking-2507","provider":"cerebras","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"gpt-oss:latest":{"mode":"chat","base_model":"gpt-oss:latest","provider":"ollama","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"atla/selene-mini:latest":{"mode":"chat","base_model":"atla/selene-mini:latest","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"atla/selene-mini:fp_16":{"mode":"chat","base_model":"atla/selene-mini:fp_16","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"Osmosis/Osmosis-Structure-0.6B:latest":{"mode":"chat","base_model":"Osmosis/Osmosis-Structure-0.6B:latest","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:1b":{"mode":"chat","base_model":"gemma3:1b","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:27b":{"mode":"chat","base_model":"gemma3:27b","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:1b-it-qat":{"mode":"chat","base_model":"gemma3:1b-it-qat","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:1b-it-q4_K_M":{"mode":"chat","base_model":"gemma3:1b-it-q4_K_M","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:1b-it-q8_0":{"mode":"chat","base_model":"gemma3:1b-it-q8_0","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:1b-it-fp16":{"mode":"chat","base_model":"gemma3:1b-it-fp16","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:4b-it-qat":{"mode":"chat","base_model":"gemma3:4b-it-qat","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:4b-it-q4_K_M":{"mode":"chat","base_model":"gemma3:4b-it-q4_K_M","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:4b-it-q8_0":{"mode":"chat","base_model":"gemma3:4b-it-q8_0","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:4b-it-fp16":{"mode":"chat","base_model":"gemma3:4b-it-fp16","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:12b-it-qat":{"mode":"chat","base_model":"gemma3:12b-it-qat","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:12b-it-q4_K_M":{"mode":"chat","base_model":"gemma3:12b-it-q4_K_M","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:12b-it-q8_0":{"mode":"chat","base_model":"gemma3:12b-it-q8_0","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:12b-it-fp16":{"mode":"chat","base_model":"gemma3:12b-it-fp16","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:27b-it-qat":{"mode":"chat","base_model":"gemma3:27b-it-qat","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:27b-it-q4_K_M":{"mode":"chat","base_model":"gemma3:27b-it-q4_K_M","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:27b-it-q8_0":{"mode":"chat","base_model":"gemma3:27b-it-q8_0","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma3:27b-it-fp16":{"mode":"chat","base_model":"gemma3:27b-it-fp16","provider":"ollama","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:0.6b-q4_K_M":{"mode":"chat","base_model":"qwen3:0.6b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:0.6b-q8_0":{"mode":"chat","base_model":"qwen3:0.6b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:0.6b-fp16":{"mode":"chat","base_model":"qwen3:0.6b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:1.7b-q4_K_M":{"mode":"chat","base_model":"qwen3:1.7b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:1.7b-q8_0":{"mode":"chat","base_model":"qwen3:1.7b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:1.7b-fp16":{"mode":"chat","base_model":"qwen3:1.7b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:4b-q4_K_M":{"mode":"chat","base_model":"qwen3:4b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:4b-q8_0":{"mode":"chat","base_model":"qwen3:4b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:4b-fp16":{"mode":"chat","base_model":"qwen3:4b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:8b-q4_K_M":{"mode":"chat","base_model":"qwen3:8b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:8b-q8_0":{"mode":"chat","base_model":"qwen3:8b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:8b-fp16":{"mode":"chat","base_model":"qwen3:8b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:14b-q4_K_M":{"mode":"chat","base_model":"qwen3:14b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:14b-q8_0":{"mode":"chat","base_model":"qwen3:14b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:14b-fp16":{"mode":"chat","base_model":"qwen3:14b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:30b-a3b-q4_K_M":{"mode":"chat","base_model":"qwen3:30b-a3b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:30b-a3b-q8_0":{"mode":"chat","base_model":"qwen3:30b-a3b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:30b-a3b-fp16":{"mode":"chat","base_model":"qwen3:30b-a3b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:32b-q4_K_M":{"mode":"chat","base_model":"qwen3:32b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:32b-q8_0":{"mode":"chat","base_model":"qwen3:32b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:32b-fp16":{"mode":"chat","base_model":"qwen3:32b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:235b-a22b-q4_K_M":{"mode":"chat","base_model":"qwen3:235b-a22b-q4_K_M","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:235b-a22b-q8_0":{"mode":"chat","base_model":"qwen3:235b-a22b-q8_0","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen3:235b-a22b-fp16":{"mode":"chat","base_model":"qwen3:235b-a22b-fp16","provider":"ollama","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:1.5b":{"mode":"chat","base_model":"deepseek-r1:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:7b":{"mode":"chat","base_model":"deepseek-r1:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:14b":{"mode":"chat","base_model":"deepseek-r1:14b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:32b":{"mode":"chat","base_model":"deepseek-r1:32b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:70b":{"mode":"chat","base_model":"deepseek-r1:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:1.5b-qwen-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:1.5b-qwen-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:1.5b-qwen-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:1.5b-qwen-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:1.5b-qwen-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:1.5b-qwen-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:14b-qwen-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:14b-qwen-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:14b-qwen-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:14b-qwen-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:14b-qwen-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:14b-qwen-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:32b-qwen-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:32b-qwen-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:32b-qwen-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:32b-qwen-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:32b-qwen-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:32b-qwen-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:671b-fp16":{"mode":"chat","base_model":"deepseek-r1:671b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:671b-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:671b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:671b-q8_0":{"mode":"chat","base_model":"deepseek-r1:671b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:70b-llama-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:70b-llama-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:70b-llama-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:70b-llama-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:70b-llama-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:70b-llama-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:7b-qwen-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:7b-qwen-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:7b-qwen-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:7b-qwen-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:7b-qwen-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:7b-qwen-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:8b-llama-distill-fp16":{"mode":"chat","base_model":"deepseek-r1:8b-llama-distill-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:8b-llama-distill-q4_K_M":{"mode":"chat","base_model":"deepseek-r1:8b-llama-distill-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"deepseek-r1:8b-llama-distill-q8_0":{"mode":"chat","base_model":"deepseek-r1:8b-llama-distill-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-fp16":{"mode":"chat","base_model":"llama3.3:70b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q2_K":{"mode":"chat","base_model":"llama3.3:70b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.3:70b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.3:70b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q4_0":{"mode":"chat","base_model":"llama3.3:70b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.3:70b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.3:70b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q5_0":{"mode":"chat","base_model":"llama3.3:70b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q5_1":{"mode":"chat","base_model":"llama3.3:70b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.3:70b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q6_K":{"mode":"chat","base_model":"llama3.3:70b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.3:70b-instruct-q8_0":{"mode":"chat","base_model":"llama3.3:70b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4:14b":{"mode":"chat","base_model":"phi4:14b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4:14b-fp16":{"mode":"chat","base_model":"phi4:14b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4:14b-q4_K_M":{"mode":"chat","base_model":"phi4:14b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4:14b-q8_0":{"mode":"chat","base_model":"phi4:14b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-fp16":{"mode":"chat","base_model":"llama3.2:1b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q2_K":{"mode":"chat","base_model":"llama3.2:1b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3.2:1b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.2:1b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.2:1b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q4_0":{"mode":"chat","base_model":"llama3.2:1b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q4_1":{"mode":"chat","base_model":"llama3.2:1b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.2:1b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.2:1b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q5_0":{"mode":"chat","base_model":"llama3.2:1b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q5_1":{"mode":"chat","base_model":"llama3.2:1b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.2:1b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3.2:1b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q6_K":{"mode":"chat","base_model":"llama3.2:1b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-instruct-q8_0":{"mode":"chat","base_model":"llama3.2:1b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-fp16":{"mode":"chat","base_model":"llama3.2:1b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q2_K":{"mode":"chat","base_model":"llama3.2:1b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q3_K_L":{"mode":"chat","base_model":"llama3.2:1b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q3_K_M":{"mode":"chat","base_model":"llama3.2:1b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q3_K_S":{"mode":"chat","base_model":"llama3.2:1b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q4_0":{"mode":"chat","base_model":"llama3.2:1b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q4_1":{"mode":"chat","base_model":"llama3.2:1b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q4_K_M":{"mode":"chat","base_model":"llama3.2:1b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q4_K_S":{"mode":"chat","base_model":"llama3.2:1b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q5_0":{"mode":"chat","base_model":"llama3.2:1b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q5_1":{"mode":"chat","base_model":"llama3.2:1b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q5_K_M":{"mode":"chat","base_model":"llama3.2:1b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q5_K_S":{"mode":"chat","base_model":"llama3.2:1b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q6_K":{"mode":"chat","base_model":"llama3.2:1b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:1b-text-q8_0":{"mode":"chat","base_model":"llama3.2:1b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-fp16":{"mode":"chat","base_model":"llama3.2:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q2_K":{"mode":"chat","base_model":"llama3.2:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3.2:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.2:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.2:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q4_0":{"mode":"chat","base_model":"llama3.2:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q4_1":{"mode":"chat","base_model":"llama3.2:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.2:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.2:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q5_0":{"mode":"chat","base_model":"llama3.2:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q5_1":{"mode":"chat","base_model":"llama3.2:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.2:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3.2:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q6_K":{"mode":"chat","base_model":"llama3.2:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-instruct-q8_0":{"mode":"chat","base_model":"llama3.2:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-fp16":{"mode":"chat","base_model":"llama3.2:3b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q2_K":{"mode":"chat","base_model":"llama3.2:3b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q3_K_L":{"mode":"chat","base_model":"llama3.2:3b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q3_K_M":{"mode":"chat","base_model":"llama3.2:3b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q3_K_S":{"mode":"chat","base_model":"llama3.2:3b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q4_0":{"mode":"chat","base_model":"llama3.2:3b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q4_1":{"mode":"chat","base_model":"llama3.2:3b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q4_K_M":{"mode":"chat","base_model":"llama3.2:3b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q4_K_S":{"mode":"chat","base_model":"llama3.2:3b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q5_0":{"mode":"chat","base_model":"llama3.2:3b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q5_1":{"mode":"chat","base_model":"llama3.2:3b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q5_K_M":{"mode":"chat","base_model":"llama3.2:3b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q5_K_S":{"mode":"chat","base_model":"llama3.2:3b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q6_K":{"mode":"chat","base_model":"llama3.2:3b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2:3b-text-q8_0":{"mode":"chat","base_model":"llama3.2:3b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-fp16":{"mode":"chat","base_model":"llama3.1:405b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q2_K":{"mode":"chat","base_model":"llama3.1:405b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3.1:405b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.1:405b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.1:405b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q4_0":{"mode":"chat","base_model":"llama3.1:405b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q4_1":{"mode":"chat","base_model":"llama3.1:405b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.1:405b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.1:405b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q5_0":{"mode":"chat","base_model":"llama3.1:405b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q5_1":{"mode":"chat","base_model":"llama3.1:405b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.1:405b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3.1:405b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q6_K":{"mode":"chat","base_model":"llama3.1:405b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-instruct-q8_0":{"mode":"chat","base_model":"llama3.1:405b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-fp16":{"mode":"chat","base_model":"llama3.1:405b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q2_K":{"mode":"chat","base_model":"llama3.1:405b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q3_K_L":{"mode":"chat","base_model":"llama3.1:405b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q3_K_M":{"mode":"chat","base_model":"llama3.1:405b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q3_K_S":{"mode":"chat","base_model":"llama3.1:405b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q4_0":{"mode":"chat","base_model":"llama3.1:405b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q4_1":{"mode":"chat","base_model":"llama3.1:405b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q4_K_M":{"mode":"chat","base_model":"llama3.1:405b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q4_K_S":{"mode":"chat","base_model":"llama3.1:405b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q5_0":{"mode":"chat","base_model":"llama3.1:405b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q5_1":{"mode":"chat","base_model":"llama3.1:405b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q5_K_M":{"mode":"chat","base_model":"llama3.1:405b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q5_K_S":{"mode":"chat","base_model":"llama3.1:405b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q6_K":{"mode":"chat","base_model":"llama3.1:405b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:405b-text-q8_0":{"mode":"chat","base_model":"llama3.1:405b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-fp16":{"mode":"chat","base_model":"llama3.1:70b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q2_K":{"mode":"chat","base_model":"llama3.1:70b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3.1:70b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.1:70b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.1:70b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q4_0":{"mode":"chat","base_model":"llama3.1:70b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.1:70b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.1:70b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q5_0":{"mode":"chat","base_model":"llama3.1:70b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q5_1":{"mode":"chat","base_model":"llama3.1:70b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.1:70b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3.1:70b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q6_K":{"mode":"chat","base_model":"llama3.1:70b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-instruct-q8_0":{"mode":"chat","base_model":"llama3.1:70b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-fp16":{"mode":"chat","base_model":"llama3.1:70b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q2_K":{"mode":"chat","base_model":"llama3.1:70b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q3_K_L":{"mode":"chat","base_model":"llama3.1:70b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q3_K_M":{"mode":"chat","base_model":"llama3.1:70b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q3_K_S":{"mode":"chat","base_model":"llama3.1:70b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q4_0":{"mode":"chat","base_model":"llama3.1:70b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q4_1":{"mode":"chat","base_model":"llama3.1:70b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q4_K_M":{"mode":"chat","base_model":"llama3.1:70b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q4_K_S":{"mode":"chat","base_model":"llama3.1:70b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q5_0":{"mode":"chat","base_model":"llama3.1:70b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q5_1":{"mode":"chat","base_model":"llama3.1:70b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q5_K_M":{"mode":"chat","base_model":"llama3.1:70b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q5_K_S":{"mode":"chat","base_model":"llama3.1:70b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q6_K":{"mode":"chat","base_model":"llama3.1:70b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:70b-text-q8_0":{"mode":"chat","base_model":"llama3.1:70b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-fp16":{"mode":"chat","base_model":"llama3.1:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q2_K":{"mode":"chat","base_model":"llama3.1:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3.1:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3.1:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3.1:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q4_0":{"mode":"chat","base_model":"llama3.1:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q4_1":{"mode":"chat","base_model":"llama3.1:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.1:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3.1:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q5_0":{"mode":"chat","base_model":"llama3.1:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q5_1":{"mode":"chat","base_model":"llama3.1:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3.1:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3.1:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q6_K":{"mode":"chat","base_model":"llama3.1:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-instruct-q8_0":{"mode":"chat","base_model":"llama3.1:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-fp16":{"mode":"chat","base_model":"llama3.1:8b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q2_K":{"mode":"chat","base_model":"llama3.1:8b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q3_K_L":{"mode":"chat","base_model":"llama3.1:8b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q3_K_M":{"mode":"chat","base_model":"llama3.1:8b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q3_K_S":{"mode":"chat","base_model":"llama3.1:8b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q4_0":{"mode":"chat","base_model":"llama3.1:8b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q4_1":{"mode":"chat","base_model":"llama3.1:8b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q4_K_M":{"mode":"chat","base_model":"llama3.1:8b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q4_K_S":{"mode":"chat","base_model":"llama3.1:8b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q5_0":{"mode":"chat","base_model":"llama3.1:8b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q5_1":{"mode":"chat","base_model":"llama3.1:8b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q5_K_M":{"mode":"chat","base_model":"llama3.1:8b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q5_K_S":{"mode":"chat","base_model":"llama3.1:8b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q6_K":{"mode":"chat","base_model":"llama3.1:8b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.1:8b-text-q8_0":{"mode":"chat","base_model":"llama3.1:8b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nomic-embed-text:137m-v1.5-fp16":{"mode":"chat","base_model":"nomic-embed-text:137m-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nomic-embed-text:v1.5":{"mode":"chat","base_model":"nomic-embed-text:v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-fp16":{"mode":"chat","base_model":"mistral:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q2_K":{"mode":"chat","base_model":"mistral:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q3_K_L":{"mode":"chat","base_model":"mistral:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q3_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q3_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q4_0":{"mode":"chat","base_model":"mistral:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q4_1":{"mode":"chat","base_model":"mistral:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q4_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q4_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q5_0":{"mode":"chat","base_model":"mistral:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q5_1":{"mode":"chat","base_model":"mistral:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q5_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q5_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q6_K":{"mode":"chat","base_model":"mistral:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-q8_0":{"mode":"chat","base_model":"mistral:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-fp16":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q2_K":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q3_K_L":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q3_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q3_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q4_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q4_1":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q4_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q4_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q5_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q5_1":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q5_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q5_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q6_K":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.2-q8_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-fp16":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q2_K":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q3_K_L":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q3_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q3_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q4_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q4_1":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q4_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q4_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q5_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q5_1":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q5_K_M":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q5_K_S":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q6_K":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-instruct-v0.3-q8_0":{"mode":"chat","base_model":"mistral:7b-instruct-v0.3-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text":{"mode":"chat","base_model":"mistral:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-fp16":{"mode":"chat","base_model":"mistral:7b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q2_K":{"mode":"chat","base_model":"mistral:7b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q3_K_L":{"mode":"chat","base_model":"mistral:7b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q3_K_M":{"mode":"chat","base_model":"mistral:7b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q3_K_S":{"mode":"chat","base_model":"mistral:7b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q4_0":{"mode":"chat","base_model":"mistral:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q4_1":{"mode":"chat","base_model":"mistral:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q4_K_M":{"mode":"chat","base_model":"mistral:7b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q4_K_S":{"mode":"chat","base_model":"mistral:7b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q5_0":{"mode":"chat","base_model":"mistral:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q5_1":{"mode":"chat","base_model":"mistral:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q5_K_M":{"mode":"chat","base_model":"mistral:7b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q5_K_S":{"mode":"chat","base_model":"mistral:7b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q6_K":{"mode":"chat","base_model":"mistral:7b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-q8_0":{"mode":"chat","base_model":"mistral:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-fp16":{"mode":"chat","base_model":"mistral:7b-text-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q2_K":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q3_K_L":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q3_K_M":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q3_K_S":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q4_0":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q4_1":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q4_K_M":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q4_K_S":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q5_0":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q5_1":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q5_K_M":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q5_K_S":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q6_K":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:7b-text-v0.2-q8_0":{"mode":"chat","base_model":"mistral:7b-text-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:instruct":{"mode":"chat","base_model":"mistral:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:text":{"mode":"chat","base_model":"mistral:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:v0.1":{"mode":"chat","base_model":"mistral:v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:v0.2":{"mode":"chat","base_model":"mistral:v0.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral:v0.3":{"mode":"chat","base_model":"mistral:v0.3","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-fp16":{"mode":"chat","base_model":"llama3:70b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q2_K":{"mode":"chat","base_model":"llama3:70b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3:70b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3:70b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3:70b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q4_0":{"mode":"chat","base_model":"llama3:70b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q4_1":{"mode":"chat","base_model":"llama3:70b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3:70b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3:70b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q5_0":{"mode":"chat","base_model":"llama3:70b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q5_1":{"mode":"chat","base_model":"llama3:70b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3:70b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3:70b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q6_K":{"mode":"chat","base_model":"llama3:70b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-instruct-q8_0":{"mode":"chat","base_model":"llama3:70b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text":{"mode":"chat","base_model":"llama3:70b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-fp16":{"mode":"chat","base_model":"llama3:70b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q2_K":{"mode":"chat","base_model":"llama3:70b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q3_K_L":{"mode":"chat","base_model":"llama3:70b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q3_K_M":{"mode":"chat","base_model":"llama3:70b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q3_K_S":{"mode":"chat","base_model":"llama3:70b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q4_0":{"mode":"chat","base_model":"llama3:70b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q4_1":{"mode":"chat","base_model":"llama3:70b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q4_K_M":{"mode":"chat","base_model":"llama3:70b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q4_K_S":{"mode":"chat","base_model":"llama3:70b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q5_0":{"mode":"chat","base_model":"llama3:70b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q5_1":{"mode":"chat","base_model":"llama3:70b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q5_K_M":{"mode":"chat","base_model":"llama3:70b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q5_K_S":{"mode":"chat","base_model":"llama3:70b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q6_K":{"mode":"chat","base_model":"llama3:70b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:70b-text-q8_0":{"mode":"chat","base_model":"llama3:70b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-fp16":{"mode":"chat","base_model":"llama3:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q2_K":{"mode":"chat","base_model":"llama3:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q3_K_L":{"mode":"chat","base_model":"llama3:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q3_K_M":{"mode":"chat","base_model":"llama3:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q3_K_S":{"mode":"chat","base_model":"llama3:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q4_0":{"mode":"chat","base_model":"llama3:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q4_1":{"mode":"chat","base_model":"llama3:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q4_K_S":{"mode":"chat","base_model":"llama3:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q5_0":{"mode":"chat","base_model":"llama3:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q5_1":{"mode":"chat","base_model":"llama3:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q5_K_M":{"mode":"chat","base_model":"llama3:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q5_K_S":{"mode":"chat","base_model":"llama3:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q6_K":{"mode":"chat","base_model":"llama3:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-instruct-q8_0":{"mode":"chat","base_model":"llama3:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text":{"mode":"chat","base_model":"llama3:8b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-fp16":{"mode":"chat","base_model":"llama3:8b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q2_K":{"mode":"chat","base_model":"llama3:8b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q3_K_L":{"mode":"chat","base_model":"llama3:8b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q3_K_M":{"mode":"chat","base_model":"llama3:8b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q3_K_S":{"mode":"chat","base_model":"llama3:8b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q4_0":{"mode":"chat","base_model":"llama3:8b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q4_1":{"mode":"chat","base_model":"llama3:8b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q4_K_M":{"mode":"chat","base_model":"llama3:8b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q4_K_S":{"mode":"chat","base_model":"llama3:8b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q5_0":{"mode":"chat","base_model":"llama3:8b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q5_1":{"mode":"chat","base_model":"llama3:8b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q5_K_M":{"mode":"chat","base_model":"llama3:8b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q5_K_S":{"mode":"chat","base_model":"llama3:8b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q6_K":{"mode":"chat","base_model":"llama3:8b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:8b-text-q8_0":{"mode":"chat","base_model":"llama3:8b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:instruct":{"mode":"chat","base_model":"llama3:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3:text":{"mode":"chat","base_model":"llama3:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b":{"mode":"chat","base_model":"qwen2.5:0.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b":{"mode":"chat","base_model":"qwen2.5:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b":{"mode":"chat","base_model":"qwen2.5:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base":{"mode":"chat","base_model":"qwen2.5:0.5b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q2_K":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q4_0":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q4_1":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q5_0":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q5_1":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-base-q8_0":{"mode":"chat","base_model":"qwen2.5:0.5b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:0.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:0.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:1.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:1.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:14b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:14b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:14b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:32b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:32b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:32b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct":{"mode":"chat","base_model":"qwen2.5:3b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:3b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:72b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:72b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:72b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5:7b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b":{"mode":"chat","base_model":"qwen:0.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b":{"mode":"chat","base_model":"qwen:1.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b":{"mode":"chat","base_model":"qwen:4b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b":{"mode":"chat","base_model":"qwen:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b":{"mode":"chat","base_model":"qwen:14b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b":{"mode":"chat","base_model":"qwen:32b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b":{"mode":"chat","base_model":"qwen:72b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b":{"mode":"chat","base_model":"qwen:110b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat":{"mode":"chat","base_model":"qwen:0.5b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:0.5b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text":{"mode":"chat","base_model":"qwen:0.5b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:0.5b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:0.5b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat":{"mode":"chat","base_model":"qwen:1.8b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-fp16":{"mode":"chat","base_model":"qwen:1.8b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q2_K":{"mode":"chat","base_model":"qwen:1.8b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q3_K_L":{"mode":"chat","base_model":"qwen:1.8b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q3_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q3_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q4_0":{"mode":"chat","base_model":"qwen:1.8b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q4_1":{"mode":"chat","base_model":"qwen:1.8b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q4_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q4_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q5_0":{"mode":"chat","base_model":"qwen:1.8b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q5_1":{"mode":"chat","base_model":"qwen:1.8b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q5_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q5_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q6_K":{"mode":"chat","base_model":"qwen:1.8b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-q8_0":{"mode":"chat","base_model":"qwen:1.8b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:1.8b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text":{"mode":"chat","base_model":"qwen:1.8b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-fp16":{"mode":"chat","base_model":"qwen:1.8b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q2_K":{"mode":"chat","base_model":"qwen:1.8b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q3_K_L":{"mode":"chat","base_model":"qwen:1.8b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q3_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q3_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q4_0":{"mode":"chat","base_model":"qwen:1.8b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q4_1":{"mode":"chat","base_model":"qwen:1.8b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q4_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q4_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q5_0":{"mode":"chat","base_model":"qwen:1.8b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q5_1":{"mode":"chat","base_model":"qwen:1.8b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q5_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q5_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q6_K":{"mode":"chat","base_model":"qwen:1.8b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-q8_0":{"mode":"chat","base_model":"qwen:1.8b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:1.8b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:1.8b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat":{"mode":"chat","base_model":"qwen:110b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:110b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:110b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:110b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:110b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat":{"mode":"chat","base_model":"qwen:14b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-fp16":{"mode":"chat","base_model":"qwen:14b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q2_K":{"mode":"chat","base_model":"qwen:14b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q3_K_L":{"mode":"chat","base_model":"qwen:14b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q3_K_M":{"mode":"chat","base_model":"qwen:14b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q3_K_S":{"mode":"chat","base_model":"qwen:14b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q4_0":{"mode":"chat","base_model":"qwen:14b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q4_1":{"mode":"chat","base_model":"qwen:14b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q4_K_M":{"mode":"chat","base_model":"qwen:14b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q4_K_S":{"mode":"chat","base_model":"qwen:14b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q5_0":{"mode":"chat","base_model":"qwen:14b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q5_1":{"mode":"chat","base_model":"qwen:14b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q5_K_M":{"mode":"chat","base_model":"qwen:14b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q5_K_S":{"mode":"chat","base_model":"qwen:14b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q6_K":{"mode":"chat","base_model":"qwen:14b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-q8_0":{"mode":"chat","base_model":"qwen:14b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:14b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text":{"mode":"chat","base_model":"qwen:14b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-fp16":{"mode":"chat","base_model":"qwen:14b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q2_K":{"mode":"chat","base_model":"qwen:14b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q3_K_L":{"mode":"chat","base_model":"qwen:14b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q3_K_M":{"mode":"chat","base_model":"qwen:14b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q3_K_S":{"mode":"chat","base_model":"qwen:14b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q4_0":{"mode":"chat","base_model":"qwen:14b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q4_1":{"mode":"chat","base_model":"qwen:14b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q4_K_M":{"mode":"chat","base_model":"qwen:14b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q4_K_S":{"mode":"chat","base_model":"qwen:14b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q5_0":{"mode":"chat","base_model":"qwen:14b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q5_1":{"mode":"chat","base_model":"qwen:14b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q5_K_M":{"mode":"chat","base_model":"qwen:14b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q5_K_S":{"mode":"chat","base_model":"qwen:14b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q6_K":{"mode":"chat","base_model":"qwen:14b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-q8_0":{"mode":"chat","base_model":"qwen:14b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:14b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:14b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:14b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat":{"mode":"chat","base_model":"qwen:32b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:32b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text":{"mode":"chat","base_model":"qwen:32b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:32b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:32b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat":{"mode":"chat","base_model":"qwen:4b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:4b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text":{"mode":"chat","base_model":"qwen:4b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:4b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:4b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:4b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat":{"mode":"chat","base_model":"qwen:72b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-fp16":{"mode":"chat","base_model":"qwen:72b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q2_K":{"mode":"chat","base_model":"qwen:72b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q3_K_L":{"mode":"chat","base_model":"qwen:72b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q3_K_M":{"mode":"chat","base_model":"qwen:72b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q3_K_S":{"mode":"chat","base_model":"qwen:72b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q4_0":{"mode":"chat","base_model":"qwen:72b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q4_1":{"mode":"chat","base_model":"qwen:72b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q4_K_M":{"mode":"chat","base_model":"qwen:72b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q4_K_S":{"mode":"chat","base_model":"qwen:72b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q5_0":{"mode":"chat","base_model":"qwen:72b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q5_1":{"mode":"chat","base_model":"qwen:72b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q5_K_M":{"mode":"chat","base_model":"qwen:72b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q5_K_S":{"mode":"chat","base_model":"qwen:72b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q6_K":{"mode":"chat","base_model":"qwen:72b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-q8_0":{"mode":"chat","base_model":"qwen:72b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:72b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text":{"mode":"chat","base_model":"qwen:72b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-fp16":{"mode":"chat","base_model":"qwen:72b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q2_K":{"mode":"chat","base_model":"qwen:72b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q3_K_L":{"mode":"chat","base_model":"qwen:72b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q3_K_M":{"mode":"chat","base_model":"qwen:72b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q3_K_S":{"mode":"chat","base_model":"qwen:72b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q4_0":{"mode":"chat","base_model":"qwen:72b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q4_1":{"mode":"chat","base_model":"qwen:72b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q4_K_M":{"mode":"chat","base_model":"qwen:72b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q4_K_S":{"mode":"chat","base_model":"qwen:72b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q5_0":{"mode":"chat","base_model":"qwen:72b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q5_1":{"mode":"chat","base_model":"qwen:72b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q5_K_M":{"mode":"chat","base_model":"qwen:72b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q5_K_S":{"mode":"chat","base_model":"qwen:72b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q6_K":{"mode":"chat","base_model":"qwen:72b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-q8_0":{"mode":"chat","base_model":"qwen:72b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:72b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:72b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:72b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat":{"mode":"chat","base_model":"qwen:7b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-fp16":{"mode":"chat","base_model":"qwen:7b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q2_K":{"mode":"chat","base_model":"qwen:7b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q3_K_L":{"mode":"chat","base_model":"qwen:7b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q3_K_M":{"mode":"chat","base_model":"qwen:7b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q3_K_S":{"mode":"chat","base_model":"qwen:7b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q4_0":{"mode":"chat","base_model":"qwen:7b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q4_1":{"mode":"chat","base_model":"qwen:7b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q4_K_M":{"mode":"chat","base_model":"qwen:7b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q4_K_S":{"mode":"chat","base_model":"qwen:7b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q5_0":{"mode":"chat","base_model":"qwen:7b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q5_1":{"mode":"chat","base_model":"qwen:7b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q5_K_M":{"mode":"chat","base_model":"qwen:7b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q5_K_S":{"mode":"chat","base_model":"qwen:7b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q6_K":{"mode":"chat","base_model":"qwen:7b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-q8_0":{"mode":"chat","base_model":"qwen:7b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-fp16":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q2_K":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q4_0":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q4_1":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q5_0":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q5_1":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q6_K":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-chat-v1.5-q8_0":{"mode":"chat","base_model":"qwen:7b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-fp16":{"mode":"chat","base_model":"qwen:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q2_K":{"mode":"chat","base_model":"qwen:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q3_K_L":{"mode":"chat","base_model":"qwen:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q3_K_M":{"mode":"chat","base_model":"qwen:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q3_K_S":{"mode":"chat","base_model":"qwen:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q4_0":{"mode":"chat","base_model":"qwen:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q4_1":{"mode":"chat","base_model":"qwen:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q4_K_M":{"mode":"chat","base_model":"qwen:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q4_K_S":{"mode":"chat","base_model":"qwen:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q5_0":{"mode":"chat","base_model":"qwen:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q5_1":{"mode":"chat","base_model":"qwen:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q5_K_M":{"mode":"chat","base_model":"qwen:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q5_K_S":{"mode":"chat","base_model":"qwen:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q6_K":{"mode":"chat","base_model":"qwen:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-q8_0":{"mode":"chat","base_model":"qwen:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text":{"mode":"chat","base_model":"qwen:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-fp16":{"mode":"chat","base_model":"qwen:7b-text-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q2_K":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q3_K_L":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q3_K_M":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q3_K_S":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q4_0":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q4_1":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q4_K_M":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q4_K_S":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q5_0":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q5_1":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q5_K_M":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q5_K_S":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q6_K":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen:7b-text-v1.5-q8_0":{"mode":"chat","base_model":"qwen:7b-text-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b":{"mode":"chat","base_model":"gemma:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct":{"mode":"chat","base_model":"gemma:2b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-fp16":{"mode":"chat","base_model":"gemma:2b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q2_K":{"mode":"chat","base_model":"gemma:2b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q3_K_L":{"mode":"chat","base_model":"gemma:2b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q3_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q3_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q4_0":{"mode":"chat","base_model":"gemma:2b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q4_1":{"mode":"chat","base_model":"gemma:2b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q4_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q4_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q5_0":{"mode":"chat","base_model":"gemma:2b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q5_1":{"mode":"chat","base_model":"gemma:2b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q5_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q5_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q6_K":{"mode":"chat","base_model":"gemma:2b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-q8_0":{"mode":"chat","base_model":"gemma:2b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-fp16":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q2_K":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q3_K_L":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q3_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q3_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q4_0":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q4_1":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q4_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q4_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q5_0":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q5_1":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q5_K_M":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q5_K_S":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q6_K":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-instruct-v1.1-q8_0":{"mode":"chat","base_model":"gemma:2b-instruct-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text":{"mode":"chat","base_model":"gemma:2b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-fp16":{"mode":"chat","base_model":"gemma:2b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q2_K":{"mode":"chat","base_model":"gemma:2b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q3_K_L":{"mode":"chat","base_model":"gemma:2b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q3_K_M":{"mode":"chat","base_model":"gemma:2b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q3_K_S":{"mode":"chat","base_model":"gemma:2b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q4_0":{"mode":"chat","base_model":"gemma:2b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q4_1":{"mode":"chat","base_model":"gemma:2b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q4_K_M":{"mode":"chat","base_model":"gemma:2b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q4_K_S":{"mode":"chat","base_model":"gemma:2b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q5_0":{"mode":"chat","base_model":"gemma:2b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q5_1":{"mode":"chat","base_model":"gemma:2b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q5_K_M":{"mode":"chat","base_model":"gemma:2b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q5_K_S":{"mode":"chat","base_model":"gemma:2b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q6_K":{"mode":"chat","base_model":"gemma:2b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-text-q8_0":{"mode":"chat","base_model":"gemma:2b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:2b-v1.1":{"mode":"chat","base_model":"gemma:2b-v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct":{"mode":"chat","base_model":"gemma:7b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-fp16":{"mode":"chat","base_model":"gemma:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q2_K":{"mode":"chat","base_model":"gemma:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q3_K_L":{"mode":"chat","base_model":"gemma:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q3_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q3_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q4_0":{"mode":"chat","base_model":"gemma:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q4_1":{"mode":"chat","base_model":"gemma:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q4_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q4_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q5_0":{"mode":"chat","base_model":"gemma:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q5_1":{"mode":"chat","base_model":"gemma:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q5_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q5_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q6_K":{"mode":"chat","base_model":"gemma:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-q8_0":{"mode":"chat","base_model":"gemma:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-fp16":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q2_K":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q3_K_L":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q3_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q3_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q4_0":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q4_1":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q4_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q4_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q5_0":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q5_1":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q5_K_M":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q5_K_S":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q6_K":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-instruct-v1.1-q8_0":{"mode":"chat","base_model":"gemma:7b-instruct-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text":{"mode":"chat","base_model":"gemma:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-fp16":{"mode":"chat","base_model":"gemma:7b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q2_K":{"mode":"chat","base_model":"gemma:7b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q3_K_L":{"mode":"chat","base_model":"gemma:7b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q3_K_M":{"mode":"chat","base_model":"gemma:7b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q3_K_S":{"mode":"chat","base_model":"gemma:7b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q4_0":{"mode":"chat","base_model":"gemma:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q4_1":{"mode":"chat","base_model":"gemma:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q4_K_M":{"mode":"chat","base_model":"gemma:7b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q4_K_S":{"mode":"chat","base_model":"gemma:7b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q5_0":{"mode":"chat","base_model":"gemma:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q5_1":{"mode":"chat","base_model":"gemma:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q5_K_M":{"mode":"chat","base_model":"gemma:7b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q5_K_S":{"mode":"chat","base_model":"gemma:7b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q6_K":{"mode":"chat","base_model":"gemma:7b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-text-q8_0":{"mode":"chat","base_model":"gemma:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:7b-v1.1":{"mode":"chat","base_model":"gemma:7b-v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:instruct":{"mode":"chat","base_model":"gemma:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:text":{"mode":"chat","base_model":"gemma:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma:v1.1":{"mode":"chat","base_model":"gemma:v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b":{"mode":"chat","base_model":"qwen2:0.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b":{"mode":"chat","base_model":"qwen2:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b":{"mode":"chat","base_model":"qwen2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b":{"mode":"chat","base_model":"qwen2:72b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct":{"mode":"chat","base_model":"qwen2:0.5b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2:0.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:0.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2:0.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct":{"mode":"chat","base_model":"qwen2:1.5b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2:1.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:1.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2:1.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-fp16":{"mode":"chat","base_model":"qwen2:72b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q2_K":{"mode":"chat","base_model":"qwen2:72b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2:72b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2:72b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2:72b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q4_0":{"mode":"chat","base_model":"qwen2:72b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q4_1":{"mode":"chat","base_model":"qwen2:72b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2:72b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2:72b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q5_0":{"mode":"chat","base_model":"qwen2:72b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q5_1":{"mode":"chat","base_model":"qwen2:72b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2:72b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2:72b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q6_K":{"mode":"chat","base_model":"qwen2:72b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-instruct-q8_0":{"mode":"chat","base_model":"qwen2:72b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text":{"mode":"chat","base_model":"qwen2:72b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-fp16":{"mode":"chat","base_model":"qwen2:72b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q2_K":{"mode":"chat","base_model":"qwen2:72b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q3_K_L":{"mode":"chat","base_model":"qwen2:72b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q3_K_M":{"mode":"chat","base_model":"qwen2:72b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q3_K_S":{"mode":"chat","base_model":"qwen2:72b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q4_0":{"mode":"chat","base_model":"qwen2:72b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q4_1":{"mode":"chat","base_model":"qwen2:72b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q4_K_M":{"mode":"chat","base_model":"qwen2:72b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q4_K_S":{"mode":"chat","base_model":"qwen2:72b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q5_0":{"mode":"chat","base_model":"qwen2:72b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q5_1":{"mode":"chat","base_model":"qwen2:72b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q5_K_M":{"mode":"chat","base_model":"qwen2:72b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q5_K_S":{"mode":"chat","base_model":"qwen2:72b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q6_K":{"mode":"chat","base_model":"qwen2:72b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:72b-text-q8_0":{"mode":"chat","base_model":"qwen2:72b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-fp16":{"mode":"chat","base_model":"qwen2:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q2_K":{"mode":"chat","base_model":"qwen2:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q4_0":{"mode":"chat","base_model":"qwen2:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q4_1":{"mode":"chat","base_model":"qwen2:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q5_0":{"mode":"chat","base_model":"qwen2:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q5_1":{"mode":"chat","base_model":"qwen2:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q6_K":{"mode":"chat","base_model":"qwen2:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-instruct-q8_0":{"mode":"chat","base_model":"qwen2:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text":{"mode":"chat","base_model":"qwen2:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q2_K":{"mode":"chat","base_model":"qwen2:7b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q3_K_L":{"mode":"chat","base_model":"qwen2:7b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q3_K_M":{"mode":"chat","base_model":"qwen2:7b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q3_K_S":{"mode":"chat","base_model":"qwen2:7b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q4_0":{"mode":"chat","base_model":"qwen2:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q4_1":{"mode":"chat","base_model":"qwen2:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q4_K_M":{"mode":"chat","base_model":"qwen2:7b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q4_K_S":{"mode":"chat","base_model":"qwen2:7b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q5_0":{"mode":"chat","base_model":"qwen2:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q5_1":{"mode":"chat","base_model":"qwen2:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2:7b-text-q8_0":{"mode":"chat","base_model":"qwen2:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:0.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:0.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:1.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:1.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base":{"mode":"chat","base_model":"qwen2.5-coder:14b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:14b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:14b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base":{"mode":"chat","base_model":"qwen2.5-coder:32b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:32b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:32b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base":{"mode":"chat","base_model":"qwen2.5-coder:3b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:3b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base":{"mode":"chat","base_model":"qwen2.5-coder:7b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-fp16":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-base-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-fp16":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q2_K":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q4_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q4_1":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q5_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q5_1":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q6_K":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2.5-coder:7b-instruct-q8_0":{"mode":"chat","base_model":"qwen2.5-coder:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b":{"mode":"chat","base_model":"llava:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b":{"mode":"chat","base_model":"llava:34b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-fp16":{"mode":"chat","base_model":"llava:13b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q2_K":{"mode":"chat","base_model":"llava:13b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q3_K_L":{"mode":"chat","base_model":"llava:13b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q3_K_M":{"mode":"chat","base_model":"llava:13b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q3_K_S":{"mode":"chat","base_model":"llava:13b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q4_0":{"mode":"chat","base_model":"llava:13b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q4_1":{"mode":"chat","base_model":"llava:13b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q4_K_M":{"mode":"chat","base_model":"llava:13b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q4_K_S":{"mode":"chat","base_model":"llava:13b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q5_0":{"mode":"chat","base_model":"llava:13b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q5_1":{"mode":"chat","base_model":"llava:13b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q5_K_M":{"mode":"chat","base_model":"llava:13b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q5_K_S":{"mode":"chat","base_model":"llava:13b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q6_K":{"mode":"chat","base_model":"llava:13b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.5-q8_0":{"mode":"chat","base_model":"llava:13b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6":{"mode":"chat","base_model":"llava:13b-v1.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-fp16":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q2_K":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q3_K_L":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q3_K_M":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q3_K_S":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q4_0":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q4_1":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q4_K_M":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q4_K_S":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q5_0":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q5_1":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q5_K_M":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q5_K_S":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q6_K":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:13b-v1.6-vicuna-q8_0":{"mode":"chat","base_model":"llava:13b-v1.6-vicuna-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6":{"mode":"chat","base_model":"llava:34b-v1.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-fp16":{"mode":"chat","base_model":"llava:34b-v1.6-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q2_K":{"mode":"chat","base_model":"llava:34b-v1.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q3_K_L":{"mode":"chat","base_model":"llava:34b-v1.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q3_K_M":{"mode":"chat","base_model":"llava:34b-v1.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q3_K_S":{"mode":"chat","base_model":"llava:34b-v1.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q4_0":{"mode":"chat","base_model":"llava:34b-v1.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q4_1":{"mode":"chat","base_model":"llava:34b-v1.6-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q4_K_M":{"mode":"chat","base_model":"llava:34b-v1.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q4_K_S":{"mode":"chat","base_model":"llava:34b-v1.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q5_0":{"mode":"chat","base_model":"llava:34b-v1.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q5_1":{"mode":"chat","base_model":"llava:34b-v1.6-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q5_K_M":{"mode":"chat","base_model":"llava:34b-v1.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q5_K_S":{"mode":"chat","base_model":"llava:34b-v1.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q6_K":{"mode":"chat","base_model":"llava:34b-v1.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:34b-v1.6-q8_0":{"mode":"chat","base_model":"llava:34b-v1.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-fp16":{"mode":"chat","base_model":"llava:7b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q2_K":{"mode":"chat","base_model":"llava:7b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q3_K_L":{"mode":"chat","base_model":"llava:7b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q3_K_M":{"mode":"chat","base_model":"llava:7b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q3_K_S":{"mode":"chat","base_model":"llava:7b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q4_0":{"mode":"chat","base_model":"llava:7b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q4_1":{"mode":"chat","base_model":"llava:7b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q4_K_M":{"mode":"chat","base_model":"llava:7b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q4_K_S":{"mode":"chat","base_model":"llava:7b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q5_0":{"mode":"chat","base_model":"llava:7b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q5_1":{"mode":"chat","base_model":"llava:7b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q5_K_M":{"mode":"chat","base_model":"llava:7b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q5_K_S":{"mode":"chat","base_model":"llava:7b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q6_K":{"mode":"chat","base_model":"llava:7b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.5-q8_0":{"mode":"chat","base_model":"llava:7b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6":{"mode":"chat","base_model":"llava:7b-v1.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-fp16":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q2_K":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q3_K_L":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q3_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q3_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q4_0":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q4_1":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q4_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q4_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q5_0":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q5_1":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q5_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q5_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q6_K":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-mistral-q8_0":{"mode":"chat","base_model":"llava:7b-v1.6-mistral-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-fp16":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q2_K":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q3_K_L":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q3_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q3_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q4_0":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q4_1":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q4_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q4_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q5_0":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q5_1":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q5_K_M":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q5_K_S":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q6_K":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:7b-v1.6-vicuna-q8_0":{"mode":"chat","base_model":"llava:7b-v1.6-vicuna-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava:v1.6":{"mode":"chat","base_model":"llava:v1.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b":{"mode":"chat","base_model":"gemma2:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b":{"mode":"chat","base_model":"gemma2:27b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-fp16":{"mode":"chat","base_model":"gemma2:27b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q2_K":{"mode":"chat","base_model":"gemma2:27b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q3_K_L":{"mode":"chat","base_model":"gemma2:27b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q3_K_M":{"mode":"chat","base_model":"gemma2:27b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q3_K_S":{"mode":"chat","base_model":"gemma2:27b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q4_0":{"mode":"chat","base_model":"gemma2:27b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q4_1":{"mode":"chat","base_model":"gemma2:27b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q4_K_M":{"mode":"chat","base_model":"gemma2:27b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q4_K_S":{"mode":"chat","base_model":"gemma2:27b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q5_0":{"mode":"chat","base_model":"gemma2:27b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q5_1":{"mode":"chat","base_model":"gemma2:27b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q5_K_M":{"mode":"chat","base_model":"gemma2:27b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q5_K_S":{"mode":"chat","base_model":"gemma2:27b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q6_K":{"mode":"chat","base_model":"gemma2:27b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-instruct-q8_0":{"mode":"chat","base_model":"gemma2:27b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-fp16":{"mode":"chat","base_model":"gemma2:27b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q2_K":{"mode":"chat","base_model":"gemma2:27b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q3_K_L":{"mode":"chat","base_model":"gemma2:27b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q3_K_M":{"mode":"chat","base_model":"gemma2:27b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q3_K_S":{"mode":"chat","base_model":"gemma2:27b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q4_0":{"mode":"chat","base_model":"gemma2:27b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q4_1":{"mode":"chat","base_model":"gemma2:27b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q4_K_M":{"mode":"chat","base_model":"gemma2:27b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q4_K_S":{"mode":"chat","base_model":"gemma2:27b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q5_0":{"mode":"chat","base_model":"gemma2:27b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q5_1":{"mode":"chat","base_model":"gemma2:27b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q5_K_M":{"mode":"chat","base_model":"gemma2:27b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q5_K_S":{"mode":"chat","base_model":"gemma2:27b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q6_K":{"mode":"chat","base_model":"gemma2:27b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:27b-text-q8_0":{"mode":"chat","base_model":"gemma2:27b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-fp16":{"mode":"chat","base_model":"gemma2:2b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q2_K":{"mode":"chat","base_model":"gemma2:2b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q3_K_L":{"mode":"chat","base_model":"gemma2:2b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q3_K_M":{"mode":"chat","base_model":"gemma2:2b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q3_K_S":{"mode":"chat","base_model":"gemma2:2b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q4_0":{"mode":"chat","base_model":"gemma2:2b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q4_1":{"mode":"chat","base_model":"gemma2:2b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q4_K_M":{"mode":"chat","base_model":"gemma2:2b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q4_K_S":{"mode":"chat","base_model":"gemma2:2b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q5_0":{"mode":"chat","base_model":"gemma2:2b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q5_1":{"mode":"chat","base_model":"gemma2:2b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q5_K_M":{"mode":"chat","base_model":"gemma2:2b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q5_K_S":{"mode":"chat","base_model":"gemma2:2b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q6_K":{"mode":"chat","base_model":"gemma2:2b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-instruct-q8_0":{"mode":"chat","base_model":"gemma2:2b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-fp16":{"mode":"chat","base_model":"gemma2:2b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q2_K":{"mode":"chat","base_model":"gemma2:2b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q3_K_L":{"mode":"chat","base_model":"gemma2:2b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q3_K_M":{"mode":"chat","base_model":"gemma2:2b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q3_K_S":{"mode":"chat","base_model":"gemma2:2b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q4_0":{"mode":"chat","base_model":"gemma2:2b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q4_1":{"mode":"chat","base_model":"gemma2:2b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q4_K_M":{"mode":"chat","base_model":"gemma2:2b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q4_K_S":{"mode":"chat","base_model":"gemma2:2b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q5_0":{"mode":"chat","base_model":"gemma2:2b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q5_1":{"mode":"chat","base_model":"gemma2:2b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q5_K_M":{"mode":"chat","base_model":"gemma2:2b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q5_K_S":{"mode":"chat","base_model":"gemma2:2b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q6_K":{"mode":"chat","base_model":"gemma2:2b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:2b-text-q8_0":{"mode":"chat","base_model":"gemma2:2b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-fp16":{"mode":"chat","base_model":"gemma2:9b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q2_K":{"mode":"chat","base_model":"gemma2:9b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q3_K_L":{"mode":"chat","base_model":"gemma2:9b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q3_K_M":{"mode":"chat","base_model":"gemma2:9b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q3_K_S":{"mode":"chat","base_model":"gemma2:9b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q4_0":{"mode":"chat","base_model":"gemma2:9b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q4_1":{"mode":"chat","base_model":"gemma2:9b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q4_K_M":{"mode":"chat","base_model":"gemma2:9b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q4_K_S":{"mode":"chat","base_model":"gemma2:9b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q5_0":{"mode":"chat","base_model":"gemma2:9b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q5_1":{"mode":"chat","base_model":"gemma2:9b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q5_K_M":{"mode":"chat","base_model":"gemma2:9b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q5_K_S":{"mode":"chat","base_model":"gemma2:9b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q6_K":{"mode":"chat","base_model":"gemma2:9b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-instruct-q8_0":{"mode":"chat","base_model":"gemma2:9b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-fp16":{"mode":"chat","base_model":"gemma2:9b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q2_K":{"mode":"chat","base_model":"gemma2:9b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q3_K_L":{"mode":"chat","base_model":"gemma2:9b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q3_K_M":{"mode":"chat","base_model":"gemma2:9b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q3_K_S":{"mode":"chat","base_model":"gemma2:9b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q4_0":{"mode":"chat","base_model":"gemma2:9b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q4_1":{"mode":"chat","base_model":"gemma2:9b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q4_K_M":{"mode":"chat","base_model":"gemma2:9b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q4_K_S":{"mode":"chat","base_model":"gemma2:9b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q5_0":{"mode":"chat","base_model":"gemma2:9b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q5_1":{"mode":"chat","base_model":"gemma2:9b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q5_K_M":{"mode":"chat","base_model":"gemma2:9b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q5_K_S":{"mode":"chat","base_model":"gemma2:9b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q6_K":{"mode":"chat","base_model":"gemma2:9b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"gemma2:9b-text-q8_0":{"mode":"chat","base_model":"gemma2:9b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-fp16":{"mode":"chat","base_model":"llama2:13b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q2_K":{"mode":"chat","base_model":"llama2:13b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q3_K_L":{"mode":"chat","base_model":"llama2:13b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q3_K_M":{"mode":"chat","base_model":"llama2:13b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q3_K_S":{"mode":"chat","base_model":"llama2:13b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q4_0":{"mode":"chat","base_model":"llama2:13b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q4_1":{"mode":"chat","base_model":"llama2:13b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q4_K_M":{"mode":"chat","base_model":"llama2:13b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q4_K_S":{"mode":"chat","base_model":"llama2:13b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q5_0":{"mode":"chat","base_model":"llama2:13b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q5_1":{"mode":"chat","base_model":"llama2:13b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q5_K_M":{"mode":"chat","base_model":"llama2:13b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q5_K_S":{"mode":"chat","base_model":"llama2:13b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q6_K":{"mode":"chat","base_model":"llama2:13b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-chat-q8_0":{"mode":"chat","base_model":"llama2:13b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text":{"mode":"chat","base_model":"llama2:13b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-fp16":{"mode":"chat","base_model":"llama2:13b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q2_K":{"mode":"chat","base_model":"llama2:13b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q3_K_L":{"mode":"chat","base_model":"llama2:13b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q3_K_M":{"mode":"chat","base_model":"llama2:13b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q3_K_S":{"mode":"chat","base_model":"llama2:13b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q4_0":{"mode":"chat","base_model":"llama2:13b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q4_1":{"mode":"chat","base_model":"llama2:13b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q4_K_M":{"mode":"chat","base_model":"llama2:13b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q4_K_S":{"mode":"chat","base_model":"llama2:13b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q5_0":{"mode":"chat","base_model":"llama2:13b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q5_1":{"mode":"chat","base_model":"llama2:13b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q5_K_M":{"mode":"chat","base_model":"llama2:13b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q5_K_S":{"mode":"chat","base_model":"llama2:13b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q6_K":{"mode":"chat","base_model":"llama2:13b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:13b-text-q8_0":{"mode":"chat","base_model":"llama2:13b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-fp16":{"mode":"chat","base_model":"llama2:70b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q2_K":{"mode":"chat","base_model":"llama2:70b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q3_K_L":{"mode":"chat","base_model":"llama2:70b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q3_K_M":{"mode":"chat","base_model":"llama2:70b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q3_K_S":{"mode":"chat","base_model":"llama2:70b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q4_0":{"mode":"chat","base_model":"llama2:70b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q4_1":{"mode":"chat","base_model":"llama2:70b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q4_K_M":{"mode":"chat","base_model":"llama2:70b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q4_K_S":{"mode":"chat","base_model":"llama2:70b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q5_0":{"mode":"chat","base_model":"llama2:70b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q5_1":{"mode":"chat","base_model":"llama2:70b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q5_K_M":{"mode":"chat","base_model":"llama2:70b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q5_K_S":{"mode":"chat","base_model":"llama2:70b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q6_K":{"mode":"chat","base_model":"llama2:70b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-chat-q8_0":{"mode":"chat","base_model":"llama2:70b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text":{"mode":"chat","base_model":"llama2:70b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-fp16":{"mode":"chat","base_model":"llama2:70b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q2_K":{"mode":"chat","base_model":"llama2:70b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q3_K_L":{"mode":"chat","base_model":"llama2:70b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q3_K_M":{"mode":"chat","base_model":"llama2:70b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q3_K_S":{"mode":"chat","base_model":"llama2:70b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q4_0":{"mode":"chat","base_model":"llama2:70b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q4_1":{"mode":"chat","base_model":"llama2:70b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q4_K_M":{"mode":"chat","base_model":"llama2:70b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q4_K_S":{"mode":"chat","base_model":"llama2:70b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q5_0":{"mode":"chat","base_model":"llama2:70b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q5_1":{"mode":"chat","base_model":"llama2:70b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q5_K_M":{"mode":"chat","base_model":"llama2:70b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q5_K_S":{"mode":"chat","base_model":"llama2:70b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q6_K":{"mode":"chat","base_model":"llama2:70b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:70b-text-q8_0":{"mode":"chat","base_model":"llama2:70b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-fp16":{"mode":"chat","base_model":"llama2:7b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q2_K":{"mode":"chat","base_model":"llama2:7b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q3_K_L":{"mode":"chat","base_model":"llama2:7b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q3_K_M":{"mode":"chat","base_model":"llama2:7b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q3_K_S":{"mode":"chat","base_model":"llama2:7b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q4_0":{"mode":"chat","base_model":"llama2:7b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q4_1":{"mode":"chat","base_model":"llama2:7b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q4_K_M":{"mode":"chat","base_model":"llama2:7b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q4_K_S":{"mode":"chat","base_model":"llama2:7b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q5_0":{"mode":"chat","base_model":"llama2:7b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q5_1":{"mode":"chat","base_model":"llama2:7b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q5_K_M":{"mode":"chat","base_model":"llama2:7b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q5_K_S":{"mode":"chat","base_model":"llama2:7b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q6_K":{"mode":"chat","base_model":"llama2:7b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-chat-q8_0":{"mode":"chat","base_model":"llama2:7b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text":{"mode":"chat","base_model":"llama2:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-fp16":{"mode":"chat","base_model":"llama2:7b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q2_K":{"mode":"chat","base_model":"llama2:7b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q3_K_L":{"mode":"chat","base_model":"llama2:7b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q3_K_M":{"mode":"chat","base_model":"llama2:7b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q3_K_S":{"mode":"chat","base_model":"llama2:7b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q4_0":{"mode":"chat","base_model":"llama2:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q4_1":{"mode":"chat","base_model":"llama2:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q4_K_M":{"mode":"chat","base_model":"llama2:7b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q4_K_S":{"mode":"chat","base_model":"llama2:7b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q5_0":{"mode":"chat","base_model":"llama2:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q5_1":{"mode":"chat","base_model":"llama2:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q5_K_M":{"mode":"chat","base_model":"llama2:7b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q5_K_S":{"mode":"chat","base_model":"llama2:7b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q6_K":{"mode":"chat","base_model":"llama2:7b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:7b-text-q8_0":{"mode":"chat","base_model":"llama2:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:chat":{"mode":"chat","base_model":"llama2:chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2:text":{"mode":"chat","base_model":"llama2:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b":{"mode":"chat","base_model":"phi3:3.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b":{"mode":"chat","base_model":"phi3:14b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-instruct":{"mode":"chat","base_model":"phi3:14b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-fp16":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q2_K":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q3_K_L":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q3_K_M":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q3_K_S":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q4_0":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q4_1":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q4_K_M":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q4_K_S":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q5_0":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q5_1":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q5_K_M":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q5_K_S":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q6_K":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-128k-instruct-q8_0":{"mode":"chat","base_model":"phi3:14b-medium-128k-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-fp16":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q2_K":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q3_K_L":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q3_K_M":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q3_K_S":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q4_0":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q4_1":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q4_K_M":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q4_K_S":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q5_0":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q5_1":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q5_K_M":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q5_K_S":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q6_K":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:14b-medium-4k-instruct-q8_0":{"mode":"chat","base_model":"phi3:14b-medium-4k-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-instruct":{"mode":"chat","base_model":"phi3:3.8b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-fp16":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q2_K":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q3_K_L":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q3_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q3_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q4_0":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q4_1":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q4_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q4_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q5_0":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q5_1":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q5_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q5_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q6_K":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-128k-instruct-q8_0":{"mode":"chat","base_model":"phi3:3.8b-mini-128k-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-fp16":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q2_K":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q3_K_L":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q3_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q3_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q4_0":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q4_1":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q4_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q4_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q5_0":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q5_1":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q5_K_M":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q5_K_S":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q6_K":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:3.8b-mini-4k-instruct-q8_0":{"mode":"chat","base_model":"phi3:3.8b-mini-4k-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3:instruct":{"mode":"chat","base_model":"phi3:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:medium":{"mode":"chat","base_model":"phi3:medium","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:medium-128k":{"mode":"chat","base_model":"phi3:medium-128k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:medium-4k":{"mode":"chat","base_model":"phi3:medium-4k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:mini":{"mode":"chat","base_model":"phi3:mini","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:mini-128k":{"mode":"chat","base_model":"phi3:mini-128k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"phi3:mini-4k":{"mode":"chat","base_model":"phi3:mini-4k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code":{"mode":"chat","base_model":"codellama:13b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-fp16":{"mode":"chat","base_model":"codellama:13b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q2_K":{"mode":"chat","base_model":"codellama:13b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q3_K_L":{"mode":"chat","base_model":"codellama:13b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q3_K_M":{"mode":"chat","base_model":"codellama:13b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q3_K_S":{"mode":"chat","base_model":"codellama:13b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q4_0":{"mode":"chat","base_model":"codellama:13b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q4_1":{"mode":"chat","base_model":"codellama:13b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q4_K_M":{"mode":"chat","base_model":"codellama:13b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q4_K_S":{"mode":"chat","base_model":"codellama:13b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q5_0":{"mode":"chat","base_model":"codellama:13b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q5_1":{"mode":"chat","base_model":"codellama:13b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q5_K_M":{"mode":"chat","base_model":"codellama:13b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q5_K_S":{"mode":"chat","base_model":"codellama:13b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q6_K":{"mode":"chat","base_model":"codellama:13b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-code-q8_0":{"mode":"chat","base_model":"codellama:13b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-fp16":{"mode":"chat","base_model":"codellama:13b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q2_K":{"mode":"chat","base_model":"codellama:13b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q3_K_L":{"mode":"chat","base_model":"codellama:13b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q3_K_M":{"mode":"chat","base_model":"codellama:13b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q3_K_S":{"mode":"chat","base_model":"codellama:13b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q4_0":{"mode":"chat","base_model":"codellama:13b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q4_1":{"mode":"chat","base_model":"codellama:13b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q4_K_M":{"mode":"chat","base_model":"codellama:13b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q4_K_S":{"mode":"chat","base_model":"codellama:13b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q5_0":{"mode":"chat","base_model":"codellama:13b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q5_1":{"mode":"chat","base_model":"codellama:13b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q5_K_M":{"mode":"chat","base_model":"codellama:13b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q5_K_S":{"mode":"chat","base_model":"codellama:13b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q6_K":{"mode":"chat","base_model":"codellama:13b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-instruct-q8_0":{"mode":"chat","base_model":"codellama:13b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-fp16":{"mode":"chat","base_model":"codellama:13b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q2_K":{"mode":"chat","base_model":"codellama:13b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q3_K_L":{"mode":"chat","base_model":"codellama:13b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q3_K_M":{"mode":"chat","base_model":"codellama:13b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q3_K_S":{"mode":"chat","base_model":"codellama:13b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q4_0":{"mode":"chat","base_model":"codellama:13b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q4_1":{"mode":"chat","base_model":"codellama:13b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q4_K_M":{"mode":"chat","base_model":"codellama:13b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q4_K_S":{"mode":"chat","base_model":"codellama:13b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q5_0":{"mode":"chat","base_model":"codellama:13b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q5_1":{"mode":"chat","base_model":"codellama:13b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q5_K_M":{"mode":"chat","base_model":"codellama:13b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q5_K_S":{"mode":"chat","base_model":"codellama:13b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q6_K":{"mode":"chat","base_model":"codellama:13b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:13b-python-q8_0":{"mode":"chat","base_model":"codellama:13b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code":{"mode":"chat","base_model":"codellama:34b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q2_K":{"mode":"chat","base_model":"codellama:34b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q3_K_L":{"mode":"chat","base_model":"codellama:34b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q3_K_M":{"mode":"chat","base_model":"codellama:34b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q3_K_S":{"mode":"chat","base_model":"codellama:34b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q4_0":{"mode":"chat","base_model":"codellama:34b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q4_1":{"mode":"chat","base_model":"codellama:34b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q4_K_M":{"mode":"chat","base_model":"codellama:34b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q4_K_S":{"mode":"chat","base_model":"codellama:34b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q5_0":{"mode":"chat","base_model":"codellama:34b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q5_1":{"mode":"chat","base_model":"codellama:34b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q5_K_M":{"mode":"chat","base_model":"codellama:34b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q5_K_S":{"mode":"chat","base_model":"codellama:34b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q6_K":{"mode":"chat","base_model":"codellama:34b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-code-q8_0":{"mode":"chat","base_model":"codellama:34b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-fp16":{"mode":"chat","base_model":"codellama:34b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q2_K":{"mode":"chat","base_model":"codellama:34b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q3_K_L":{"mode":"chat","base_model":"codellama:34b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q3_K_M":{"mode":"chat","base_model":"codellama:34b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q3_K_S":{"mode":"chat","base_model":"codellama:34b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q4_0":{"mode":"chat","base_model":"codellama:34b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q4_1":{"mode":"chat","base_model":"codellama:34b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q4_K_M":{"mode":"chat","base_model":"codellama:34b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q4_K_S":{"mode":"chat","base_model":"codellama:34b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q5_0":{"mode":"chat","base_model":"codellama:34b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q5_1":{"mode":"chat","base_model":"codellama:34b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q5_K_M":{"mode":"chat","base_model":"codellama:34b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q5_K_S":{"mode":"chat","base_model":"codellama:34b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q6_K":{"mode":"chat","base_model":"codellama:34b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-instruct-q8_0":{"mode":"chat","base_model":"codellama:34b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-fp16":{"mode":"chat","base_model":"codellama:34b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q2_K":{"mode":"chat","base_model":"codellama:34b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q3_K_L":{"mode":"chat","base_model":"codellama:34b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q3_K_M":{"mode":"chat","base_model":"codellama:34b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q3_K_S":{"mode":"chat","base_model":"codellama:34b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q4_0":{"mode":"chat","base_model":"codellama:34b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q4_1":{"mode":"chat","base_model":"codellama:34b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q4_K_M":{"mode":"chat","base_model":"codellama:34b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q4_K_S":{"mode":"chat","base_model":"codellama:34b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q5_0":{"mode":"chat","base_model":"codellama:34b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q5_1":{"mode":"chat","base_model":"codellama:34b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q5_K_M":{"mode":"chat","base_model":"codellama:34b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q5_K_S":{"mode":"chat","base_model":"codellama:34b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q6_K":{"mode":"chat","base_model":"codellama:34b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:34b-python-q8_0":{"mode":"chat","base_model":"codellama:34b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code":{"mode":"chat","base_model":"codellama:70b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-fp16":{"mode":"chat","base_model":"codellama:70b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q2_K":{"mode":"chat","base_model":"codellama:70b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q3_K_L":{"mode":"chat","base_model":"codellama:70b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q3_K_M":{"mode":"chat","base_model":"codellama:70b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q3_K_S":{"mode":"chat","base_model":"codellama:70b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q4_0":{"mode":"chat","base_model":"codellama:70b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q4_1":{"mode":"chat","base_model":"codellama:70b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q4_K_M":{"mode":"chat","base_model":"codellama:70b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q4_K_S":{"mode":"chat","base_model":"codellama:70b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q5_0":{"mode":"chat","base_model":"codellama:70b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q5_1":{"mode":"chat","base_model":"codellama:70b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q5_K_M":{"mode":"chat","base_model":"codellama:70b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q5_K_S":{"mode":"chat","base_model":"codellama:70b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q6_K":{"mode":"chat","base_model":"codellama:70b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-code-q8_0":{"mode":"chat","base_model":"codellama:70b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-fp16":{"mode":"chat","base_model":"codellama:70b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q2_K":{"mode":"chat","base_model":"codellama:70b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q3_K_L":{"mode":"chat","base_model":"codellama:70b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q3_K_M":{"mode":"chat","base_model":"codellama:70b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q3_K_S":{"mode":"chat","base_model":"codellama:70b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q4_0":{"mode":"chat","base_model":"codellama:70b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q4_1":{"mode":"chat","base_model":"codellama:70b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q4_K_M":{"mode":"chat","base_model":"codellama:70b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q4_K_S":{"mode":"chat","base_model":"codellama:70b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q5_0":{"mode":"chat","base_model":"codellama:70b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q5_1":{"mode":"chat","base_model":"codellama:70b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q5_K_M":{"mode":"chat","base_model":"codellama:70b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q5_K_S":{"mode":"chat","base_model":"codellama:70b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q6_K":{"mode":"chat","base_model":"codellama:70b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-instruct-q8_0":{"mode":"chat","base_model":"codellama:70b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-fp16":{"mode":"chat","base_model":"codellama:70b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q2_K":{"mode":"chat","base_model":"codellama:70b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q3_K_L":{"mode":"chat","base_model":"codellama:70b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q3_K_M":{"mode":"chat","base_model":"codellama:70b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q3_K_S":{"mode":"chat","base_model":"codellama:70b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q4_0":{"mode":"chat","base_model":"codellama:70b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q4_1":{"mode":"chat","base_model":"codellama:70b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q4_K_M":{"mode":"chat","base_model":"codellama:70b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q4_K_S":{"mode":"chat","base_model":"codellama:70b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q5_0":{"mode":"chat","base_model":"codellama:70b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q5_1":{"mode":"chat","base_model":"codellama:70b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q5_K_M":{"mode":"chat","base_model":"codellama:70b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q5_K_S":{"mode":"chat","base_model":"codellama:70b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q6_K":{"mode":"chat","base_model":"codellama:70b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:70b-python-q8_0":{"mode":"chat","base_model":"codellama:70b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code":{"mode":"chat","base_model":"codellama:7b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-fp16":{"mode":"chat","base_model":"codellama:7b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q2_K":{"mode":"chat","base_model":"codellama:7b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q3_K_L":{"mode":"chat","base_model":"codellama:7b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q3_K_M":{"mode":"chat","base_model":"codellama:7b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q3_K_S":{"mode":"chat","base_model":"codellama:7b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q4_0":{"mode":"chat","base_model":"codellama:7b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q4_1":{"mode":"chat","base_model":"codellama:7b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q4_K_M":{"mode":"chat","base_model":"codellama:7b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q4_K_S":{"mode":"chat","base_model":"codellama:7b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q5_0":{"mode":"chat","base_model":"codellama:7b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q5_1":{"mode":"chat","base_model":"codellama:7b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q5_K_M":{"mode":"chat","base_model":"codellama:7b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q5_K_S":{"mode":"chat","base_model":"codellama:7b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q6_K":{"mode":"chat","base_model":"codellama:7b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-code-q8_0":{"mode":"chat","base_model":"codellama:7b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-fp16":{"mode":"chat","base_model":"codellama:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q2_K":{"mode":"chat","base_model":"codellama:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q3_K_L":{"mode":"chat","base_model":"codellama:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q3_K_M":{"mode":"chat","base_model":"codellama:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q3_K_S":{"mode":"chat","base_model":"codellama:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q4_0":{"mode":"chat","base_model":"codellama:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q4_1":{"mode":"chat","base_model":"codellama:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q4_K_M":{"mode":"chat","base_model":"codellama:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q4_K_S":{"mode":"chat","base_model":"codellama:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q5_0":{"mode":"chat","base_model":"codellama:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q5_1":{"mode":"chat","base_model":"codellama:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q5_K_M":{"mode":"chat","base_model":"codellama:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q5_K_S":{"mode":"chat","base_model":"codellama:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q6_K":{"mode":"chat","base_model":"codellama:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-instruct-q8_0":{"mode":"chat","base_model":"codellama:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-fp16":{"mode":"chat","base_model":"codellama:7b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q2_K":{"mode":"chat","base_model":"codellama:7b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q3_K_L":{"mode":"chat","base_model":"codellama:7b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q3_K_M":{"mode":"chat","base_model":"codellama:7b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q3_K_S":{"mode":"chat","base_model":"codellama:7b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q4_0":{"mode":"chat","base_model":"codellama:7b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q4_1":{"mode":"chat","base_model":"codellama:7b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q4_K_M":{"mode":"chat","base_model":"codellama:7b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q4_K_S":{"mode":"chat","base_model":"codellama:7b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q5_0":{"mode":"chat","base_model":"codellama:7b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q5_1":{"mode":"chat","base_model":"codellama:7b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q5_K_M":{"mode":"chat","base_model":"codellama:7b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q5_K_S":{"mode":"chat","base_model":"codellama:7b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q6_K":{"mode":"chat","base_model":"codellama:7b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:7b-python-q8_0":{"mode":"chat","base_model":"codellama:7b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:code":{"mode":"chat","base_model":"codellama:code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"codellama:instruct":{"mode":"chat","base_model":"codellama:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codellama:python":{"mode":"chat","base_model":"codellama:python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mxbai-embed-large:335m":{"mode":"chat","base_model":"mxbai-embed-large:335m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mxbai-embed-large:335m-v1-fp16":{"mode":"chat","base_model":"mxbai-embed-large:335m-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mxbai-embed-large:v1":{"mode":"chat","base_model":"mxbai-embed-large:v1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:11b":{"mode":"chat","base_model":"llama3.2-vision:11b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:90b":{"mode":"chat","base_model":"llama3.2-vision:90b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:11b-instruct-fp16":{"mode":"chat","base_model":"llama3.2-vision:11b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:11b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.2-vision:11b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:11b-instruct-q8_0":{"mode":"chat","base_model":"llama3.2-vision:11b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:90b-instruct-fp16":{"mode":"chat","base_model":"llama3.2-vision:90b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:90b-instruct-q4_K_M":{"mode":"chat","base_model":"llama3.2-vision:90b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"llama3.2-vision:90b-instruct-q8_0":{"mode":"chat","base_model":"llama3.2-vision:90b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b":{"mode":"chat","base_model":"tinyllama:1.1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat":{"mode":"chat","base_model":"tinyllama:1.1b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-fp16":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q2_K":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q3_K_L":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q3_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q3_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q4_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q4_1":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q4_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q4_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q5_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q5_1":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q5_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q5_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q6_K":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v0.6-q8_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v0.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-fp16":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q2_K":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q3_K_L":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q3_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q3_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q4_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q4_1":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q4_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q4_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q5_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q5_1":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q5_K_M":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q5_K_S":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q6_K":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:1.1b-chat-v1-q8_0":{"mode":"chat","base_model":"tinyllama:1.1b-chat-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:chat":{"mode":"chat","base_model":"tinyllama:chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:v0.6":{"mode":"chat","base_model":"tinyllama:v0.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinyllama:v1":{"mode":"chat","base_model":"tinyllama:v1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b":{"mode":"chat","base_model":"mistral-nemo:12b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-fp16":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q2_K":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q3_K_L":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q3_K_M":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q3_K_S":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q4_0":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q4_1":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q4_K_M":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q4_K_S":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q5_0":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q5_1":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q5_K_M":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q5_K_S":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q6_K":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-nemo:12b-instruct-2407-q8_0":{"mode":"chat","base_model":"mistral-nemo:12b-instruct-2407-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-fp16":{"mode":"chat","base_model":"starcoder2:15b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct":{"mode":"chat","base_model":"starcoder2:15b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-q4_0":{"mode":"chat","base_model":"starcoder2:15b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-fp16":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q2_K":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q3_K_L":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q3_K_M":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q3_K_S":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q4_0":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q4_1":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q4_K_M":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q4_K_S":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q5_0":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q5_1":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q5_K_M":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q5_K_S":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q6_K":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-instruct-v0.1-q8_0":{"mode":"chat","base_model":"starcoder2:15b-instruct-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q2_K":{"mode":"chat","base_model":"starcoder2:15b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q3_K_L":{"mode":"chat","base_model":"starcoder2:15b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q3_K_M":{"mode":"chat","base_model":"starcoder2:15b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q3_K_S":{"mode":"chat","base_model":"starcoder2:15b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q4_0":{"mode":"chat","base_model":"starcoder2:15b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q4_1":{"mode":"chat","base_model":"starcoder2:15b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q4_K_M":{"mode":"chat","base_model":"starcoder2:15b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q4_K_S":{"mode":"chat","base_model":"starcoder2:15b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q5_0":{"mode":"chat","base_model":"starcoder2:15b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q5_1":{"mode":"chat","base_model":"starcoder2:15b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q5_K_M":{"mode":"chat","base_model":"starcoder2:15b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q5_K_S":{"mode":"chat","base_model":"starcoder2:15b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q6_K":{"mode":"chat","base_model":"starcoder2:15b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:15b-q8_0":{"mode":"chat","base_model":"starcoder2:15b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-fp16":{"mode":"chat","base_model":"starcoder2:3b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q2_K":{"mode":"chat","base_model":"starcoder2:3b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q3_K_L":{"mode":"chat","base_model":"starcoder2:3b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q3_K_M":{"mode":"chat","base_model":"starcoder2:3b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q3_K_S":{"mode":"chat","base_model":"starcoder2:3b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q4_0":{"mode":"chat","base_model":"starcoder2:3b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q4_1":{"mode":"chat","base_model":"starcoder2:3b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q4_K_M":{"mode":"chat","base_model":"starcoder2:3b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q4_K_S":{"mode":"chat","base_model":"starcoder2:3b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q5_0":{"mode":"chat","base_model":"starcoder2:3b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q5_1":{"mode":"chat","base_model":"starcoder2:3b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q5_K_M":{"mode":"chat","base_model":"starcoder2:3b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q5_K_S":{"mode":"chat","base_model":"starcoder2:3b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q6_K":{"mode":"chat","base_model":"starcoder2:3b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:3b-q8_0":{"mode":"chat","base_model":"starcoder2:3b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-fp16":{"mode":"chat","base_model":"starcoder2:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q2_K":{"mode":"chat","base_model":"starcoder2:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q3_K_L":{"mode":"chat","base_model":"starcoder2:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q3_K_M":{"mode":"chat","base_model":"starcoder2:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q3_K_S":{"mode":"chat","base_model":"starcoder2:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q4_0":{"mode":"chat","base_model":"starcoder2:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q4_1":{"mode":"chat","base_model":"starcoder2:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q4_K_M":{"mode":"chat","base_model":"starcoder2:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q4_K_S":{"mode":"chat","base_model":"starcoder2:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q5_0":{"mode":"chat","base_model":"starcoder2:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q5_1":{"mode":"chat","base_model":"starcoder2:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q5_K_M":{"mode":"chat","base_model":"starcoder2:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q5_K_S":{"mode":"chat","base_model":"starcoder2:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q6_K":{"mode":"chat","base_model":"starcoder2:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:7b-q8_0":{"mode":"chat","base_model":"starcoder2:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder2:instruct":{"mode":"chat","base_model":"starcoder2:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b":{"mode":"chat","base_model":"deepseek-coder-v2:16b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b":{"mode":"chat","base_model":"deepseek-coder-v2:236b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-fp16":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q2_K":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q3_K_L":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q3_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q3_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q4_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q4_1":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q4_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q4_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q5_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q5_1":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q5_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q5_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q6_K":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-base-q8_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-fp16":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q2_K":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q3_K_L":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q3_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q3_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q4_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q4_1":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q4_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q4_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q5_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q5_1":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q5_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q5_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q6_K":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:16b-lite-instruct-q8_0":{"mode":"chat","base_model":"deepseek-coder-v2:16b-lite-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-fp16":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q2_K":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q4_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q4_1":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q5_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q5_1":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q6_K":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-base-q8_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-fp16":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q2_K":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q3_K_L":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q3_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q3_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q4_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q4_1":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q4_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q4_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q5_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q5_1":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q5_K_M":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q5_K_S":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q6_K":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:236b-instruct-q8_0":{"mode":"chat","base_model":"deepseek-coder-v2:236b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder-v2:lite":{"mode":"chat","base_model":"deepseek-coder-v2:lite","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:22m":{"mode":"chat","base_model":"snowflake-arctic-embed:22m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:33m":{"mode":"chat","base_model":"snowflake-arctic-embed:33m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:110m":{"mode":"chat","base_model":"snowflake-arctic-embed:110m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:137m":{"mode":"chat","base_model":"snowflake-arctic-embed:137m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:335m":{"mode":"chat","base_model":"snowflake-arctic-embed:335m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:110m-m-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed:110m-m-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:137m-m-long-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed:137m-m-long-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:22m-xs-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed:22m-xs-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:335m-l-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed:335m-l-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:33m-s-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed:33m-s-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:l":{"mode":"chat","base_model":"snowflake-arctic-embed:l","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:m":{"mode":"chat","base_model":"snowflake-arctic-embed:m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:m-long":{"mode":"chat","base_model":"snowflake-arctic-embed:m-long","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:s":{"mode":"chat","base_model":"snowflake-arctic-embed:s","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed:xs":{"mode":"chat","base_model":"snowflake-arctic-embed:xs","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v3:671b":{"mode":"chat","base_model":"deepseek-v3:671b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v3:671b-fp16":{"mode":"chat","base_model":"deepseek-v3:671b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v3:671b-q4_K_M":{"mode":"chat","base_model":"deepseek-v3:671b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v3:671b-q8_0":{"mode":"chat","base_model":"deepseek-v3:671b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b":{"mode":"chat","base_model":"llama2-uncensored:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b":{"mode":"chat","base_model":"llama2-uncensored:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat":{"mode":"chat","base_model":"llama2-uncensored:70b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q2_K":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q3_K_L":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q3_K_M":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q3_K_S":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q4_0":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q4_1":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q4_K_M":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q4_K_S":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q5_0":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q5_1":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q5_K_M":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q5_K_S":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q6_K":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:70b-chat-q8_0":{"mode":"chat","base_model":"llama2-uncensored:70b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat":{"mode":"chat","base_model":"llama2-uncensored:7b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-fp16":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q2_K":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q3_K_L":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q3_K_M":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q3_K_S":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q4_0":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q4_1":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q4_K_M":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q4_K_S":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q5_0":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q5_1":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q5_K_M":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q5_K_S":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q6_K":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-uncensored:7b-chat-q8_0":{"mode":"chat","base_model":"llama2-uncensored:7b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b":{"mode":"chat","base_model":"deepseek-coder:1.3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b":{"mode":"chat","base_model":"deepseek-coder:33b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base":{"mode":"chat","base_model":"deepseek-coder:1.3b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-fp16":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q2_K":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q4_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q4_1":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q5_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q5_1":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q6_K":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-base-q8_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-fp16":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q2_K":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q4_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q4_1":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q5_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q5_1":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q6_K":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:1.3b-instruct-q8_0":{"mode":"chat","base_model":"deepseek-coder:1.3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base":{"mode":"chat","base_model":"deepseek-coder:33b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-fp16":{"mode":"chat","base_model":"deepseek-coder:33b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q2_K":{"mode":"chat","base_model":"deepseek-coder:33b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:33b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q4_0":{"mode":"chat","base_model":"deepseek-coder:33b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q4_1":{"mode":"chat","base_model":"deepseek-coder:33b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q5_0":{"mode":"chat","base_model":"deepseek-coder:33b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q5_1":{"mode":"chat","base_model":"deepseek-coder:33b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q6_K":{"mode":"chat","base_model":"deepseek-coder:33b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-base-q8_0":{"mode":"chat","base_model":"deepseek-coder:33b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-fp16":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q2_K":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q4_0":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q4_1":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q5_0":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q5_1":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q6_K":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:33b-instruct-q8_0":{"mode":"chat","base_model":"deepseek-coder:33b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base":{"mode":"chat","base_model":"deepseek-coder:6.7b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-fp16":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q2_K":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q4_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q4_1":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q5_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q5_1":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q6_K":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-base-q8_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-fp16":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q2_K":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q3_K_L":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q3_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q3_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q4_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q4_1":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q4_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q4_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q5_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q5_1":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q5_K_M":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q5_K_S":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q6_K":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:6.7b-instruct-q8_0":{"mode":"chat","base_model":"deepseek-coder:6.7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:base":{"mode":"chat","base_model":"deepseek-coder:base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-coder:instruct":{"mode":"chat","base_model":"deepseek-coder:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-fp16":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q2_K":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q3_K_L":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q3_K_M":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q3_K_S":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q4_0":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q4_1":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q4_K_M":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q4_K_S":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q5_0":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q5_1":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q5_K_M":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q5_K_S":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q6_K":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-instruct-v0.1-q8_0":{"mode":"chat","base_model":"mixtral:8x22b-instruct-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text":{"mode":"chat","base_model":"mixtral:8x22b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-fp16":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q2_K":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q3_K_L":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q3_K_M":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q3_K_S":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q4_0":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q4_1":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q4_K_M":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q4_K_S":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q5_0":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q5_1":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q5_K_M":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q5_K_S":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q6_K":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x22b-text-v0.1-q8_0":{"mode":"chat","base_model":"mixtral:8x22b-text-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-fp16":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q2_K":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q3_K_L":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q3_K_M":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q3_K_S":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q4_0":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q4_1":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q4_K_M":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q4_K_S":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q5_0":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q5_1":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q5_K_M":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q5_K_S":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q6_K":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-instruct-v0.1-q8_0":{"mode":"chat","base_model":"mixtral:8x7b-instruct-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text":{"mode":"chat","base_model":"mixtral:8x7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-fp16":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q2_K":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q3_K_L":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q3_K_M":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q3_K_S":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q4_0":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q4_1":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q4_K_M":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q4_K_S":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q5_0":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q5_1":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q5_K_M":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q5_K_S":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q6_K":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:8x7b-text-v0.1-q8_0":{"mode":"chat","base_model":"mixtral:8x7b-text-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:instruct":{"mode":"chat","base_model":"mixtral:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:text":{"mode":"chat","base_model":"mixtral:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:v0.1":{"mode":"chat","base_model":"mixtral:v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mixtral:v0.1-instruct":{"mode":"chat","base_model":"mixtral:v0.1-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b":{"mode":"chat","base_model":"dolphin-mixtral:8x7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b":{"mode":"chat","base_model":"dolphin-mixtral:8x22b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-fp16":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q2_K":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q3_K_L":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q3_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q3_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q4_0":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q4_1":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q4_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q4_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q5_0":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q5_1":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q5_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q5_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q6_K":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x22b-v2.9-q8_0":{"mode":"chat","base_model":"dolphin-mixtral:8x22b-v2.9-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-fp16":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q2_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q3_K_L":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q3_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q3_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q4_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q4_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q4_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q4_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q5_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q5_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q5_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q5_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q6_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.5-q8_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-fp16":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q2_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q3_K_L":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q3_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q3_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q4_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q4_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q4_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q4_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q5_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q5_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q5_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q5_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q6_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.6-q8_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-fp16":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q2_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q3_K_L":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q3_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q3_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q4_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q4_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q4_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q4_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q5_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q5_1":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q5_K_M":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q5_K_S":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q6_K":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:8x7b-v2.7-q8_0":{"mode":"chat","base_model":"dolphin-mixtral:8x7b-v2.7-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:v2.5":{"mode":"chat","base_model":"dolphin-mixtral:v2.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:v2.6":{"mode":"chat","base_model":"dolphin-mixtral:v2.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mixtral:v2.7":{"mode":"chat","base_model":"dolphin-mixtral:v2.7","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code":{"mode":"chat","base_model":"codegemma:2b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-fp16":{"mode":"chat","base_model":"codegemma:2b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q2_K":{"mode":"chat","base_model":"codegemma:2b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q3_K_L":{"mode":"chat","base_model":"codegemma:2b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q3_K_M":{"mode":"chat","base_model":"codegemma:2b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q3_K_S":{"mode":"chat","base_model":"codegemma:2b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q4_0":{"mode":"chat","base_model":"codegemma:2b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q4_1":{"mode":"chat","base_model":"codegemma:2b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q4_K_M":{"mode":"chat","base_model":"codegemma:2b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q4_K_S":{"mode":"chat","base_model":"codegemma:2b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q5_0":{"mode":"chat","base_model":"codegemma:2b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q5_1":{"mode":"chat","base_model":"codegemma:2b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q5_K_M":{"mode":"chat","base_model":"codegemma:2b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q5_K_S":{"mode":"chat","base_model":"codegemma:2b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q6_K":{"mode":"chat","base_model":"codegemma:2b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-q8_0":{"mode":"chat","base_model":"codegemma:2b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-fp16":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q2_K":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q3_K_L":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q3_K_M":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q3_K_S":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q4_0":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q4_1":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q4_K_M":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q4_K_S":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q5_0":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q5_1":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q5_K_M":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q5_K_S":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q6_K":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-code-v1.1-q8_0":{"mode":"chat","base_model":"codegemma:2b-code-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:2b-v1.1":{"mode":"chat","base_model":"codegemma:2b-v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code":{"mode":"chat","base_model":"codegemma:7b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-fp16":{"mode":"chat","base_model":"codegemma:7b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q2_K":{"mode":"chat","base_model":"codegemma:7b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q3_K_L":{"mode":"chat","base_model":"codegemma:7b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q3_K_M":{"mode":"chat","base_model":"codegemma:7b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q3_K_S":{"mode":"chat","base_model":"codegemma:7b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q4_0":{"mode":"chat","base_model":"codegemma:7b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q4_1":{"mode":"chat","base_model":"codegemma:7b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q4_K_M":{"mode":"chat","base_model":"codegemma:7b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q4_K_S":{"mode":"chat","base_model":"codegemma:7b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q5_0":{"mode":"chat","base_model":"codegemma:7b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q5_1":{"mode":"chat","base_model":"codegemma:7b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q5_K_M":{"mode":"chat","base_model":"codegemma:7b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q5_K_S":{"mode":"chat","base_model":"codegemma:7b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q6_K":{"mode":"chat","base_model":"codegemma:7b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-code-q8_0":{"mode":"chat","base_model":"codegemma:7b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct":{"mode":"chat","base_model":"codegemma:7b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-fp16":{"mode":"chat","base_model":"codegemma:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q2_K":{"mode":"chat","base_model":"codegemma:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q3_K_L":{"mode":"chat","base_model":"codegemma:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q3_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q3_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q4_0":{"mode":"chat","base_model":"codegemma:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q4_1":{"mode":"chat","base_model":"codegemma:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q4_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q4_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q5_0":{"mode":"chat","base_model":"codegemma:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q5_1":{"mode":"chat","base_model":"codegemma:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q5_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q5_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q6_K":{"mode":"chat","base_model":"codegemma:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-q8_0":{"mode":"chat","base_model":"codegemma:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-fp16":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q2_K":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q3_K_L":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q3_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q3_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q4_0":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q4_1":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q4_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q4_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q5_0":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q5_1":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q5_K_M":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q5_K_S":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q6_K":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-instruct-v1.1-q8_0":{"mode":"chat","base_model":"codegemma:7b-instruct-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:7b-v1.1":{"mode":"chat","base_model":"codegemma:7b-v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegemma:code":{"mode":"chat","base_model":"codegemma:code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"codegemma:instruct":{"mode":"chat","base_model":"codegemma:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:32b":{"mode":"chat","base_model":"openthinker:32b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:32b-fp16":{"mode":"chat","base_model":"openthinker:32b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:32b-q4_K_M":{"mode":"chat","base_model":"openthinker:32b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:32b-q8_0":{"mode":"chat","base_model":"openthinker:32b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:7b-fp16":{"mode":"chat","base_model":"openthinker:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:7b-q4_K_M":{"mode":"chat","base_model":"openthinker:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openthinker:7b-q8_0":{"mode":"chat","base_model":"openthinker:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b":{"mode":"chat","base_model":"phi:2.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-fp16":{"mode":"chat","base_model":"phi:2.7b-chat-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q2_K":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q3_K_L":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q3_K_M":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q3_K_S":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q4_0":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q4_1":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q4_K_M":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q4_K_S":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q5_0":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q5_1":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q5_K_M":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q5_K_S":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q6_K":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:2.7b-chat-v2-q8_0":{"mode":"chat","base_model":"phi:2.7b-chat-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi:chat":{"mode":"chat","base_model":"phi:chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bge-m3:567m":{"mode":"chat","base_model":"bge-m3:567m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bge-m3:567m-fp16":{"mode":"chat","base_model":"bge-m3:567m-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b":{"mode":"chat","base_model":"minicpm-v:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-fp16":{"mode":"chat","base_model":"minicpm-v:8b-2.6-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q2_K":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q3_K_L":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q3_K_M":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q3_K_S":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q4_0":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q4_1":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q4_K_M":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q4_K_S":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q5_0":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q5_1":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q5_K_M":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q5_K_S":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q6_K":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"minicpm-v:8b-2.6-q8_0":{"mode":"chat","base_model":"minicpm-v:8b-2.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-llama3:8b":{"mode":"chat","base_model":"llava-llama3:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-llama3:8b-v1.1-fp16":{"mode":"chat","base_model":"llava-llama3:8b-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-llama3:8b-v1.1-q4_0":{"mode":"chat","base_model":"llava-llama3:8b-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b":{"mode":"chat","base_model":"wizardlm2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-fp16":{"mode":"chat","base_model":"wizardlm2:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q2_K":{"mode":"chat","base_model":"wizardlm2:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q3_K_L":{"mode":"chat","base_model":"wizardlm2:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q3_K_M":{"mode":"chat","base_model":"wizardlm2:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q3_K_S":{"mode":"chat","base_model":"wizardlm2:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q4_0":{"mode":"chat","base_model":"wizardlm2:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q4_1":{"mode":"chat","base_model":"wizardlm2:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q4_K_M":{"mode":"chat","base_model":"wizardlm2:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q4_K_S":{"mode":"chat","base_model":"wizardlm2:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q5_0":{"mode":"chat","base_model":"wizardlm2:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q5_1":{"mode":"chat","base_model":"wizardlm2:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q5_K_M":{"mode":"chat","base_model":"wizardlm2:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q5_K_S":{"mode":"chat","base_model":"wizardlm2:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q6_K":{"mode":"chat","base_model":"wizardlm2:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:7b-q8_0":{"mode":"chat","base_model":"wizardlm2:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:8x22b-fp16":{"mode":"chat","base_model":"wizardlm2:8x22b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:8x22b-q2_K":{"mode":"chat","base_model":"wizardlm2:8x22b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:8x22b-q4_0":{"mode":"chat","base_model":"wizardlm2:8x22b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm2:8x22b-q8_0":{"mode":"chat","base_model":"wizardlm2:8x22b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b":{"mode":"chat","base_model":"dolphin-mistral:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2":{"mode":"chat","base_model":"dolphin-mistral:7b-v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.1-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.2.1-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.2.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-dpo-laser-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-dpo-laser-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.6-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-fp16":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q2_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q3_K_L":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q3_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q3_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q4_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q4_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q4_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q4_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q5_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q5_1":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q5_K_M":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q5_K_S":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q6_K":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:7b-v2.8-q8_0":{"mode":"chat","base_model":"dolphin-mistral:7b-v2.8-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2":{"mode":"chat","base_model":"dolphin-mistral:v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2.1":{"mode":"chat","base_model":"dolphin-mistral:v2.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2.2":{"mode":"chat","base_model":"dolphin-mistral:v2.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2.2.1":{"mode":"chat","base_model":"dolphin-mistral:v2.2.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2.6":{"mode":"chat","base_model":"dolphin-mistral:v2.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-mistral:v2.8":{"mode":"chat","base_model":"dolphin-mistral:v2.8","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:22m":{"mode":"chat","base_model":"all-minilm:22m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:33m":{"mode":"chat","base_model":"all-minilm:33m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:22m-l6-v2-fp16":{"mode":"chat","base_model":"all-minilm:22m-l6-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:33m-l12-v2-fp16":{"mode":"chat","base_model":"all-minilm:33m-l12-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:l12":{"mode":"chat","base_model":"all-minilm:l12","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:l12-v2":{"mode":"chat","base_model":"all-minilm:l12-v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:l6":{"mode":"chat","base_model":"all-minilm:l6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:l6-v2":{"mode":"chat","base_model":"all-minilm:l6-v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"all-minilm:v2":{"mode":"chat","base_model":"all-minilm:v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m":{"mode":"chat","base_model":"smollm2:135m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m":{"mode":"chat","base_model":"smollm2:360m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b":{"mode":"chat","base_model":"smollm2:1.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-fp16":{"mode":"chat","base_model":"smollm2:1.7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q2_K":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q3_K_L":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q3_K_M":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q3_K_S":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q4_0":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q4_1":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q4_K_M":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q4_K_S":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q5_0":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q5_1":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q5_K_M":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q5_K_S":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q6_K":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:1.7b-instruct-q8_0":{"mode":"chat","base_model":"smollm2:1.7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-fp16":{"mode":"chat","base_model":"smollm2:135m-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q2_K":{"mode":"chat","base_model":"smollm2:135m-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q3_K_L":{"mode":"chat","base_model":"smollm2:135m-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q3_K_M":{"mode":"chat","base_model":"smollm2:135m-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q3_K_S":{"mode":"chat","base_model":"smollm2:135m-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q4_0":{"mode":"chat","base_model":"smollm2:135m-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q4_1":{"mode":"chat","base_model":"smollm2:135m-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q4_K_M":{"mode":"chat","base_model":"smollm2:135m-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q4_K_S":{"mode":"chat","base_model":"smollm2:135m-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q5_0":{"mode":"chat","base_model":"smollm2:135m-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q5_1":{"mode":"chat","base_model":"smollm2:135m-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q5_K_M":{"mode":"chat","base_model":"smollm2:135m-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q5_K_S":{"mode":"chat","base_model":"smollm2:135m-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q6_K":{"mode":"chat","base_model":"smollm2:135m-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:135m-instruct-q8_0":{"mode":"chat","base_model":"smollm2:135m-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-fp16":{"mode":"chat","base_model":"smollm2:360m-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q2_K":{"mode":"chat","base_model":"smollm2:360m-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q3_K_L":{"mode":"chat","base_model":"smollm2:360m-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q3_K_M":{"mode":"chat","base_model":"smollm2:360m-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q3_K_S":{"mode":"chat","base_model":"smollm2:360m-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q4_0":{"mode":"chat","base_model":"smollm2:360m-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q4_1":{"mode":"chat","base_model":"smollm2:360m-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q4_K_M":{"mode":"chat","base_model":"smollm2:360m-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q4_K_S":{"mode":"chat","base_model":"smollm2:360m-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q5_0":{"mode":"chat","base_model":"smollm2:360m-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q5_1":{"mode":"chat","base_model":"smollm2:360m-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q5_K_M":{"mode":"chat","base_model":"smollm2:360m-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q5_K_S":{"mode":"chat","base_model":"smollm2:360m-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q6_K":{"mode":"chat","base_model":"smollm2:360m-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm2:360m-instruct-q8_0":{"mode":"chat","base_model":"smollm2:360m-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b":{"mode":"chat","base_model":"dolphin-llama3:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b":{"mode":"chat","base_model":"dolphin-llama3:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-fp16":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q2_K":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q3_K_L":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q3_K_M":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q3_K_S":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q4_0":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q4_1":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q4_K_M":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q4_K_S":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q5_0":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q5_1":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q5_K_M":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q5_K_S":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q6_K":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:70b-v2.9-q8_0":{"mode":"chat","base_model":"dolphin-llama3:70b-v2.9-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k":{"mode":"chat","base_model":"dolphin-llama3:8b-256k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-fp16":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q2_K":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q3_K_L":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q3_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q3_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q4_0":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q4_1":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q4_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q4_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q5_0":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q5_1":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q5_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q5_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q6_K":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-256k-v2.9-q8_0":{"mode":"chat","base_model":"dolphin-llama3:8b-256k-v2.9-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-fp16":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q2_K":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q3_K_L":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q3_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q3_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q4_0":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q4_1":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q4_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q4_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q5_0":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q5_1":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q5_K_M":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q5_K_S":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q6_K":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:8b-v2.9-q8_0":{"mode":"chat","base_model":"dolphin-llama3:8b-v2.9-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-llama3:v2.9":{"mode":"chat","base_model":"dolphin-llama3:v2.9","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b":{"mode":"chat","base_model":"command-r:35b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-fp16":{"mode":"chat","base_model":"command-r:35b-08-2024-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q2_K":{"mode":"chat","base_model":"command-r:35b-08-2024-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q3_K_L":{"mode":"chat","base_model":"command-r:35b-08-2024-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q3_K_M":{"mode":"chat","base_model":"command-r:35b-08-2024-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q3_K_S":{"mode":"chat","base_model":"command-r:35b-08-2024-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q4_0":{"mode":"chat","base_model":"command-r:35b-08-2024-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q4_1":{"mode":"chat","base_model":"command-r:35b-08-2024-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q4_K_M":{"mode":"chat","base_model":"command-r:35b-08-2024-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q4_K_S":{"mode":"chat","base_model":"command-r:35b-08-2024-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q5_0":{"mode":"chat","base_model":"command-r:35b-08-2024-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q5_1":{"mode":"chat","base_model":"command-r:35b-08-2024-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q5_K_M":{"mode":"chat","base_model":"command-r:35b-08-2024-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q5_K_S":{"mode":"chat","base_model":"command-r:35b-08-2024-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q6_K":{"mode":"chat","base_model":"command-r:35b-08-2024-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-08-2024-q8_0":{"mode":"chat","base_model":"command-r:35b-08-2024-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-fp16":{"mode":"chat","base_model":"command-r:35b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q2_K":{"mode":"chat","base_model":"command-r:35b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q3_K_L":{"mode":"chat","base_model":"command-r:35b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q3_K_M":{"mode":"chat","base_model":"command-r:35b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q3_K_S":{"mode":"chat","base_model":"command-r:35b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q4_0":{"mode":"chat","base_model":"command-r:35b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q4_1":{"mode":"chat","base_model":"command-r:35b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q4_K_M":{"mode":"chat","base_model":"command-r:35b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q4_K_S":{"mode":"chat","base_model":"command-r:35b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q5_1":{"mode":"chat","base_model":"command-r:35b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q5_K_M":{"mode":"chat","base_model":"command-r:35b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q5_K_S":{"mode":"chat","base_model":"command-r:35b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q6_K":{"mode":"chat","base_model":"command-r:35b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:35b-v0.1-q8_0":{"mode":"chat","base_model":"command-r:35b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r:v0.1":{"mode":"chat","base_model":"command-r:v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b":{"mode":"chat","base_model":"orca-mini:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b":{"mode":"chat","base_model":"orca-mini:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b":{"mode":"chat","base_model":"orca-mini:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b":{"mode":"chat","base_model":"orca-mini:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-fp16":{"mode":"chat","base_model":"orca-mini:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q2_K":{"mode":"chat","base_model":"orca-mini:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q3_K_L":{"mode":"chat","base_model":"orca-mini:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q3_K_M":{"mode":"chat","base_model":"orca-mini:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q3_K_S":{"mode":"chat","base_model":"orca-mini:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q4_0":{"mode":"chat","base_model":"orca-mini:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q4_1":{"mode":"chat","base_model":"orca-mini:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q4_K_M":{"mode":"chat","base_model":"orca-mini:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q4_K_S":{"mode":"chat","base_model":"orca-mini:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q5_0":{"mode":"chat","base_model":"orca-mini:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q5_1":{"mode":"chat","base_model":"orca-mini:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q5_K_M":{"mode":"chat","base_model":"orca-mini:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q5_K_S":{"mode":"chat","base_model":"orca-mini:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q6_K":{"mode":"chat","base_model":"orca-mini:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-q8_0":{"mode":"chat","base_model":"orca-mini:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-fp16":{"mode":"chat","base_model":"orca-mini:13b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q2_K":{"mode":"chat","base_model":"orca-mini:13b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q3_K_L":{"mode":"chat","base_model":"orca-mini:13b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q3_K_M":{"mode":"chat","base_model":"orca-mini:13b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q3_K_S":{"mode":"chat","base_model":"orca-mini:13b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q4_0":{"mode":"chat","base_model":"orca-mini:13b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q4_1":{"mode":"chat","base_model":"orca-mini:13b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q4_K_M":{"mode":"chat","base_model":"orca-mini:13b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q4_K_S":{"mode":"chat","base_model":"orca-mini:13b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q5_0":{"mode":"chat","base_model":"orca-mini:13b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q5_1":{"mode":"chat","base_model":"orca-mini:13b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q5_K_M":{"mode":"chat","base_model":"orca-mini:13b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q5_K_S":{"mode":"chat","base_model":"orca-mini:13b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q6_K":{"mode":"chat","base_model":"orca-mini:13b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v2-q8_0":{"mode":"chat","base_model":"orca-mini:13b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3":{"mode":"chat","base_model":"orca-mini:13b-v3","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-fp16":{"mode":"chat","base_model":"orca-mini:13b-v3-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q2_K":{"mode":"chat","base_model":"orca-mini:13b-v3-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q3_K_L":{"mode":"chat","base_model":"orca-mini:13b-v3-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q3_K_M":{"mode":"chat","base_model":"orca-mini:13b-v3-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q3_K_S":{"mode":"chat","base_model":"orca-mini:13b-v3-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q4_0":{"mode":"chat","base_model":"orca-mini:13b-v3-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q4_1":{"mode":"chat","base_model":"orca-mini:13b-v3-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q4_K_M":{"mode":"chat","base_model":"orca-mini:13b-v3-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q4_K_S":{"mode":"chat","base_model":"orca-mini:13b-v3-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q5_0":{"mode":"chat","base_model":"orca-mini:13b-v3-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q5_1":{"mode":"chat","base_model":"orca-mini:13b-v3-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q5_K_M":{"mode":"chat","base_model":"orca-mini:13b-v3-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q5_K_S":{"mode":"chat","base_model":"orca-mini:13b-v3-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q6_K":{"mode":"chat","base_model":"orca-mini:13b-v3-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:13b-v3-q8_0":{"mode":"chat","base_model":"orca-mini:13b-v3-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-fp16":{"mode":"chat","base_model":"orca-mini:3b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-q4_0":{"mode":"chat","base_model":"orca-mini:3b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-q4_1":{"mode":"chat","base_model":"orca-mini:3b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-q5_0":{"mode":"chat","base_model":"orca-mini:3b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-q5_1":{"mode":"chat","base_model":"orca-mini:3b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:3b-q8_0":{"mode":"chat","base_model":"orca-mini:3b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3":{"mode":"chat","base_model":"orca-mini:70b-v3","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-fp16":{"mode":"chat","base_model":"orca-mini:70b-v3-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q2_K":{"mode":"chat","base_model":"orca-mini:70b-v3-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q3_K_L":{"mode":"chat","base_model":"orca-mini:70b-v3-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q3_K_M":{"mode":"chat","base_model":"orca-mini:70b-v3-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q3_K_S":{"mode":"chat","base_model":"orca-mini:70b-v3-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q4_0":{"mode":"chat","base_model":"orca-mini:70b-v3-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q4_1":{"mode":"chat","base_model":"orca-mini:70b-v3-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q4_K_M":{"mode":"chat","base_model":"orca-mini:70b-v3-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q4_K_S":{"mode":"chat","base_model":"orca-mini:70b-v3-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q5_0":{"mode":"chat","base_model":"orca-mini:70b-v3-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q5_1":{"mode":"chat","base_model":"orca-mini:70b-v3-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q5_K_M":{"mode":"chat","base_model":"orca-mini:70b-v3-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q5_K_S":{"mode":"chat","base_model":"orca-mini:70b-v3-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q6_K":{"mode":"chat","base_model":"orca-mini:70b-v3-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:70b-v3-q8_0":{"mode":"chat","base_model":"orca-mini:70b-v3-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-fp16":{"mode":"chat","base_model":"orca-mini:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q2_K":{"mode":"chat","base_model":"orca-mini:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q3_K_L":{"mode":"chat","base_model":"orca-mini:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q3_K_M":{"mode":"chat","base_model":"orca-mini:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q3_K_S":{"mode":"chat","base_model":"orca-mini:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q4_0":{"mode":"chat","base_model":"orca-mini:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q4_1":{"mode":"chat","base_model":"orca-mini:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q4_K_M":{"mode":"chat","base_model":"orca-mini:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q4_K_S":{"mode":"chat","base_model":"orca-mini:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q5_0":{"mode":"chat","base_model":"orca-mini:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q5_1":{"mode":"chat","base_model":"orca-mini:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q5_K_M":{"mode":"chat","base_model":"orca-mini:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q5_K_S":{"mode":"chat","base_model":"orca-mini:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q6_K":{"mode":"chat","base_model":"orca-mini:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-q8_0":{"mode":"chat","base_model":"orca-mini:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-fp16":{"mode":"chat","base_model":"orca-mini:7b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q2_K":{"mode":"chat","base_model":"orca-mini:7b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q3_K_L":{"mode":"chat","base_model":"orca-mini:7b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q3_K_M":{"mode":"chat","base_model":"orca-mini:7b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q3_K_S":{"mode":"chat","base_model":"orca-mini:7b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q4_0":{"mode":"chat","base_model":"orca-mini:7b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q4_1":{"mode":"chat","base_model":"orca-mini:7b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q4_K_M":{"mode":"chat","base_model":"orca-mini:7b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q4_K_S":{"mode":"chat","base_model":"orca-mini:7b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q5_0":{"mode":"chat","base_model":"orca-mini:7b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q5_1":{"mode":"chat","base_model":"orca-mini:7b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q5_K_M":{"mode":"chat","base_model":"orca-mini:7b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q5_K_S":{"mode":"chat","base_model":"orca-mini:7b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q6_K":{"mode":"chat","base_model":"orca-mini:7b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v2-q8_0":{"mode":"chat","base_model":"orca-mini:7b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3":{"mode":"chat","base_model":"orca-mini:7b-v3","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-fp16":{"mode":"chat","base_model":"orca-mini:7b-v3-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q2_K":{"mode":"chat","base_model":"orca-mini:7b-v3-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q3_K_L":{"mode":"chat","base_model":"orca-mini:7b-v3-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q3_K_M":{"mode":"chat","base_model":"orca-mini:7b-v3-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q3_K_S":{"mode":"chat","base_model":"orca-mini:7b-v3-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q4_0":{"mode":"chat","base_model":"orca-mini:7b-v3-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q4_1":{"mode":"chat","base_model":"orca-mini:7b-v3-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q4_K_M":{"mode":"chat","base_model":"orca-mini:7b-v3-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q4_K_S":{"mode":"chat","base_model":"orca-mini:7b-v3-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q5_0":{"mode":"chat","base_model":"orca-mini:7b-v3-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q5_1":{"mode":"chat","base_model":"orca-mini:7b-v3-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q5_K_M":{"mode":"chat","base_model":"orca-mini:7b-v3-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q5_K_S":{"mode":"chat","base_model":"orca-mini:7b-v3-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q6_K":{"mode":"chat","base_model":"orca-mini:7b-v3-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca-mini:7b-v3-q8_0":{"mode":"chat","base_model":"orca-mini:7b-v3-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b":{"mode":"chat","base_model":"yi:9b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-fp16":{"mode":"chat","base_model":"yi:34b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q2_K":{"mode":"chat","base_model":"yi:34b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q3_K_L":{"mode":"chat","base_model":"yi:34b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q3_K_M":{"mode":"chat","base_model":"yi:34b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q3_K_S":{"mode":"chat","base_model":"yi:34b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q4_0":{"mode":"chat","base_model":"yi:34b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q4_1":{"mode":"chat","base_model":"yi:34b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q4_K_M":{"mode":"chat","base_model":"yi:34b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q4_K_S":{"mode":"chat","base_model":"yi:34b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q5_0":{"mode":"chat","base_model":"yi:34b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q5_1":{"mode":"chat","base_model":"yi:34b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q5_K_M":{"mode":"chat","base_model":"yi:34b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q5_K_S":{"mode":"chat","base_model":"yi:34b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q6_K":{"mode":"chat","base_model":"yi:34b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-q8_0":{"mode":"chat","base_model":"yi:34b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-fp16":{"mode":"chat","base_model":"yi:34b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q2_K":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q4_0":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q4_1":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q5_0":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q5_1":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q6_K":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-chat-v1.5-q8_0":{"mode":"chat","base_model":"yi:34b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q2_K":{"mode":"chat","base_model":"yi:34b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q3_K_L":{"mode":"chat","base_model":"yi:34b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q3_K_M":{"mode":"chat","base_model":"yi:34b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q3_K_S":{"mode":"chat","base_model":"yi:34b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q4_0":{"mode":"chat","base_model":"yi:34b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q4_1":{"mode":"chat","base_model":"yi:34b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q4_K_M":{"mode":"chat","base_model":"yi:34b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q4_K_S":{"mode":"chat","base_model":"yi:34b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q5_0":{"mode":"chat","base_model":"yi:34b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q5_1":{"mode":"chat","base_model":"yi:34b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q5_K_S":{"mode":"chat","base_model":"yi:34b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-q6_K":{"mode":"chat","base_model":"yi:34b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5":{"mode":"chat","base_model":"yi:34b-v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-fp16":{"mode":"chat","base_model":"yi:34b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q2_K":{"mode":"chat","base_model":"yi:34b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:34b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:34b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:34b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q4_0":{"mode":"chat","base_model":"yi:34b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q4_1":{"mode":"chat","base_model":"yi:34b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:34b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:34b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q5_0":{"mode":"chat","base_model":"yi:34b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q5_1":{"mode":"chat","base_model":"yi:34b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:34b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:34b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q6_K":{"mode":"chat","base_model":"yi:34b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:34b-v1.5-q8_0":{"mode":"chat","base_model":"yi:34b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k":{"mode":"chat","base_model":"yi:6b-200k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-fp16":{"mode":"chat","base_model":"yi:6b-200k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q2_K":{"mode":"chat","base_model":"yi:6b-200k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q3_K_L":{"mode":"chat","base_model":"yi:6b-200k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q3_K_M":{"mode":"chat","base_model":"yi:6b-200k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q3_K_S":{"mode":"chat","base_model":"yi:6b-200k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q4_0":{"mode":"chat","base_model":"yi:6b-200k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q4_1":{"mode":"chat","base_model":"yi:6b-200k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q4_K_M":{"mode":"chat","base_model":"yi:6b-200k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q4_K_S":{"mode":"chat","base_model":"yi:6b-200k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q5_0":{"mode":"chat","base_model":"yi:6b-200k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q5_1":{"mode":"chat","base_model":"yi:6b-200k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q5_K_M":{"mode":"chat","base_model":"yi:6b-200k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q5_K_S":{"mode":"chat","base_model":"yi:6b-200k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q6_K":{"mode":"chat","base_model":"yi:6b-200k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-200k-q8_0":{"mode":"chat","base_model":"yi:6b-200k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat":{"mode":"chat","base_model":"yi:6b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-fp16":{"mode":"chat","base_model":"yi:6b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q2_K":{"mode":"chat","base_model":"yi:6b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q3_K_L":{"mode":"chat","base_model":"yi:6b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q3_K_M":{"mode":"chat","base_model":"yi:6b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q3_K_S":{"mode":"chat","base_model":"yi:6b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q4_0":{"mode":"chat","base_model":"yi:6b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q4_1":{"mode":"chat","base_model":"yi:6b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q4_K_M":{"mode":"chat","base_model":"yi:6b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q4_K_S":{"mode":"chat","base_model":"yi:6b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q5_0":{"mode":"chat","base_model":"yi:6b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q5_1":{"mode":"chat","base_model":"yi:6b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q5_K_M":{"mode":"chat","base_model":"yi:6b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q5_K_S":{"mode":"chat","base_model":"yi:6b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q6_K":{"mode":"chat","base_model":"yi:6b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-q8_0":{"mode":"chat","base_model":"yi:6b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-fp16":{"mode":"chat","base_model":"yi:6b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q2_K":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q4_0":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q4_1":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q5_0":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q5_1":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q6_K":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-chat-v1.5-q8_0":{"mode":"chat","base_model":"yi:6b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-fp16":{"mode":"chat","base_model":"yi:6b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q2_K":{"mode":"chat","base_model":"yi:6b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q3_K_L":{"mode":"chat","base_model":"yi:6b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q3_K_M":{"mode":"chat","base_model":"yi:6b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q3_K_S":{"mode":"chat","base_model":"yi:6b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q4_0":{"mode":"chat","base_model":"yi:6b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q4_1":{"mode":"chat","base_model":"yi:6b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q4_K_M":{"mode":"chat","base_model":"yi:6b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q4_K_S":{"mode":"chat","base_model":"yi:6b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q5_0":{"mode":"chat","base_model":"yi:6b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q5_1":{"mode":"chat","base_model":"yi:6b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q5_K_M":{"mode":"chat","base_model":"yi:6b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q5_K_S":{"mode":"chat","base_model":"yi:6b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q6_K":{"mode":"chat","base_model":"yi:6b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-q8_0":{"mode":"chat","base_model":"yi:6b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5":{"mode":"chat","base_model":"yi:6b-v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-fp16":{"mode":"chat","base_model":"yi:6b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q2_K":{"mode":"chat","base_model":"yi:6b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:6b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:6b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:6b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q4_0":{"mode":"chat","base_model":"yi:6b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q4_1":{"mode":"chat","base_model":"yi:6b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:6b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:6b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q5_0":{"mode":"chat","base_model":"yi:6b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q5_1":{"mode":"chat","base_model":"yi:6b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:6b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:6b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q6_K":{"mode":"chat","base_model":"yi:6b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:6b-v1.5-q8_0":{"mode":"chat","base_model":"yi:6b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat":{"mode":"chat","base_model":"yi:9b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-fp16":{"mode":"chat","base_model":"yi:9b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q2_K":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q4_0":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q4_1":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q5_0":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q5_1":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q6_K":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-chat-v1.5-q8_0":{"mode":"chat","base_model":"yi:9b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5":{"mode":"chat","base_model":"yi:9b-v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-fp16":{"mode":"chat","base_model":"yi:9b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q2_K":{"mode":"chat","base_model":"yi:9b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q3_K_L":{"mode":"chat","base_model":"yi:9b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q3_K_M":{"mode":"chat","base_model":"yi:9b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q3_K_S":{"mode":"chat","base_model":"yi:9b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q4_0":{"mode":"chat","base_model":"yi:9b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q4_1":{"mode":"chat","base_model":"yi:9b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q4_K_M":{"mode":"chat","base_model":"yi:9b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q4_K_S":{"mode":"chat","base_model":"yi:9b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q5_0":{"mode":"chat","base_model":"yi:9b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q5_1":{"mode":"chat","base_model":"yi:9b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q5_K_M":{"mode":"chat","base_model":"yi:9b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q5_K_S":{"mode":"chat","base_model":"yi:9b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q6_K":{"mode":"chat","base_model":"yi:9b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:9b-v1.5-q8_0":{"mode":"chat","base_model":"yi:9b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi:v1.5":{"mode":"chat","base_model":"yi:v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b":{"mode":"chat","base_model":"hermes3:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-fp16":{"mode":"chat","base_model":"hermes3:3b-llama3.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q2_K":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q3_K_L":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q3_K_M":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q3_K_S":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q4_0":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q4_1":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q4_K_M":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q4_K_S":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q5_0":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q5_1":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q5_K_M":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q5_K_S":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q6_K":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:3b-llama3.2-q8_0":{"mode":"chat","base_model":"hermes3:3b-llama3.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-fp16":{"mode":"chat","base_model":"hermes3:405b-llama3.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q2_K":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q3_K_L":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q3_K_M":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q3_K_S":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q4_0":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q4_1":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q4_K_M":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q4_K_S":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q5_0":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q5_1":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q5_K_M":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q5_K_S":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q6_K":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:405b-llama3.1-q8_0":{"mode":"chat","base_model":"hermes3:405b-llama3.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-fp16":{"mode":"chat","base_model":"hermes3:70b-llama3.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q2_K":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q3_K_L":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q3_K_M":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q3_K_S":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q4_0":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q4_1":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q4_K_M":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q4_K_S":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q5_0":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q5_1":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q5_K_M":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q5_K_S":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q6_K":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:70b-llama3.1-q8_0":{"mode":"chat","base_model":"hermes3:70b-llama3.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-fp16":{"mode":"chat","base_model":"hermes3:8b-llama3.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q2_K":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q3_K_L":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q3_K_M":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q3_K_S":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q4_0":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q4_1":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q4_K_M":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q4_K_S":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q5_0":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q5_1":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q5_K_M":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q5_K_S":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q6_K":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"hermes3:8b-llama3.1-q8_0":{"mode":"chat","base_model":"hermes3:8b-llama3.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin3:8b-llama3.1-fp16":{"mode":"chat","base_model":"dolphin3:8b-llama3.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin3:8b-llama3.1-q4_K_M":{"mode":"chat","base_model":"dolphin3:8b-llama3.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin3:8b-llama3.1-q8_0":{"mode":"chat","base_model":"dolphin3:8b-llama3.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b":{"mode":"chat","base_model":"phi3.5:3.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-fp16":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q2_K":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q3_K_L":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q3_K_M":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q3_K_S":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q4_0":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q4_1":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q4_K_M":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q4_K_S":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q5_0":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q5_1":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q5_K_M":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q5_K_S":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q6_K":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi3.5:3.8b-mini-instruct-q8_0":{"mode":"chat","base_model":"phi3.5:3.8b-mini-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b":{"mode":"chat","base_model":"zephyr:141b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b-v0.1":{"mode":"chat","base_model":"zephyr:141b-v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b-v0.1-fp16":{"mode":"chat","base_model":"zephyr:141b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b-v0.1-q2_K":{"mode":"chat","base_model":"zephyr:141b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b-v0.1-q4_0":{"mode":"chat","base_model":"zephyr:141b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:141b-v0.1-q8_0":{"mode":"chat","base_model":"zephyr:141b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha":{"mode":"chat","base_model":"zephyr:7b-alpha","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-fp16":{"mode":"chat","base_model":"zephyr:7b-alpha-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q2_K":{"mode":"chat","base_model":"zephyr:7b-alpha-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q3_K_L":{"mode":"chat","base_model":"zephyr:7b-alpha-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q3_K_M":{"mode":"chat","base_model":"zephyr:7b-alpha-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q3_K_S":{"mode":"chat","base_model":"zephyr:7b-alpha-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q4_0":{"mode":"chat","base_model":"zephyr:7b-alpha-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q4_1":{"mode":"chat","base_model":"zephyr:7b-alpha-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q4_K_M":{"mode":"chat","base_model":"zephyr:7b-alpha-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q4_K_S":{"mode":"chat","base_model":"zephyr:7b-alpha-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q5_0":{"mode":"chat","base_model":"zephyr:7b-alpha-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q5_1":{"mode":"chat","base_model":"zephyr:7b-alpha-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q5_K_M":{"mode":"chat","base_model":"zephyr:7b-alpha-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q5_K_S":{"mode":"chat","base_model":"zephyr:7b-alpha-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q6_K":{"mode":"chat","base_model":"zephyr:7b-alpha-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-alpha-q8_0":{"mode":"chat","base_model":"zephyr:7b-alpha-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta":{"mode":"chat","base_model":"zephyr:7b-beta","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-fp16":{"mode":"chat","base_model":"zephyr:7b-beta-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q2_K":{"mode":"chat","base_model":"zephyr:7b-beta-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q3_K_L":{"mode":"chat","base_model":"zephyr:7b-beta-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q3_K_M":{"mode":"chat","base_model":"zephyr:7b-beta-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q3_K_S":{"mode":"chat","base_model":"zephyr:7b-beta-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q4_0":{"mode":"chat","base_model":"zephyr:7b-beta-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q4_1":{"mode":"chat","base_model":"zephyr:7b-beta-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q4_K_M":{"mode":"chat","base_model":"zephyr:7b-beta-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q4_K_S":{"mode":"chat","base_model":"zephyr:7b-beta-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q5_0":{"mode":"chat","base_model":"zephyr:7b-beta-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q5_1":{"mode":"chat","base_model":"zephyr:7b-beta-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q5_K_M":{"mode":"chat","base_model":"zephyr:7b-beta-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q5_K_S":{"mode":"chat","base_model":"zephyr:7b-beta-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q6_K":{"mode":"chat","base_model":"zephyr:7b-beta-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"zephyr:7b-beta-q8_0":{"mode":"chat","base_model":"zephyr:7b-beta-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:7b":{"mode":"chat","base_model":"olmo2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:13b":{"mode":"chat","base_model":"olmo2:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:13b-1124-instruct-fp16":{"mode":"chat","base_model":"olmo2:13b-1124-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:13b-1124-instruct-q4_K_M":{"mode":"chat","base_model":"olmo2:13b-1124-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:13b-1124-instruct-q8_0":{"mode":"chat","base_model":"olmo2:13b-1124-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:7b-1124-instruct-fp16":{"mode":"chat","base_model":"olmo2:7b-1124-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:7b-1124-instruct-q4_K_M":{"mode":"chat","base_model":"olmo2:7b-1124-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"olmo2:7b-1124-instruct-q8_0":{"mode":"chat","base_model":"olmo2:7b-1124-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b":{"mode":"chat","base_model":"mistral-small:22b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:24b":{"mode":"chat","base_model":"mistral-small:24b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-fp16":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q2_K":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q3_K_L":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q3_K_M":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q3_K_S":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q4_0":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q4_1":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q4_K_M":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q4_K_S":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q5_0":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q5_1":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q5_K_M":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q5_K_S":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q6_K":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:22b-instruct-2409-q8_0":{"mode":"chat","base_model":"mistral-small:22b-instruct-2409-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:24b-instruct-2501-fp16":{"mode":"chat","base_model":"mistral-small:24b-instruct-2501-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:24b-instruct-2501-q4_K_M":{"mode":"chat","base_model":"mistral-small:24b-instruct-2501-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-small:24b-instruct-2501-q8_0":{"mode":"chat","base_model":"mistral-small:24b-instruct-2501-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b":{"mode":"chat","base_model":"codestral:22b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q2_K":{"mode":"chat","base_model":"codestral:22b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q3_K_L":{"mode":"chat","base_model":"codestral:22b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q3_K_M":{"mode":"chat","base_model":"codestral:22b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q3_K_S":{"mode":"chat","base_model":"codestral:22b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q4_0":{"mode":"chat","base_model":"codestral:22b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q4_1":{"mode":"chat","base_model":"codestral:22b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q4_K_M":{"mode":"chat","base_model":"codestral:22b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q4_K_S":{"mode":"chat","base_model":"codestral:22b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q5_0":{"mode":"chat","base_model":"codestral:22b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q5_1":{"mode":"chat","base_model":"codestral:22b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q5_K_M":{"mode":"chat","base_model":"codestral:22b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q5_K_S":{"mode":"chat","base_model":"codestral:22b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q6_K":{"mode":"chat","base_model":"codestral:22b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:22b-v0.1-q8_0":{"mode":"chat","base_model":"codestral:22b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codestral:v0.1":{"mode":"chat","base_model":"codestral:v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b":{"mode":"chat","base_model":"granite-code:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b":{"mode":"chat","base_model":"granite-code:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b":{"mode":"chat","base_model":"granite-code:20b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b":{"mode":"chat","base_model":"granite-code:34b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base":{"mode":"chat","base_model":"granite-code:20b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-fp16":{"mode":"chat","base_model":"granite-code:20b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q2_K":{"mode":"chat","base_model":"granite-code:20b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q3_K_L":{"mode":"chat","base_model":"granite-code:20b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q3_K_M":{"mode":"chat","base_model":"granite-code:20b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q3_K_S":{"mode":"chat","base_model":"granite-code:20b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q4_0":{"mode":"chat","base_model":"granite-code:20b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q4_1":{"mode":"chat","base_model":"granite-code:20b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q4_K_M":{"mode":"chat","base_model":"granite-code:20b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q4_K_S":{"mode":"chat","base_model":"granite-code:20b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q5_0":{"mode":"chat","base_model":"granite-code:20b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q5_1":{"mode":"chat","base_model":"granite-code:20b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q5_K_M":{"mode":"chat","base_model":"granite-code:20b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q5_K_S":{"mode":"chat","base_model":"granite-code:20b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q6_K":{"mode":"chat","base_model":"granite-code:20b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-base-q8_0":{"mode":"chat","base_model":"granite-code:20b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct":{"mode":"chat","base_model":"granite-code:20b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-fp16":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q2_K":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q3_K_L":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q3_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q3_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q4_0":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q4_1":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q4_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q4_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q5_0":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q5_1":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q5_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q5_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q6_K":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-8k-q8_0":{"mode":"chat","base_model":"granite-code:20b-instruct-8k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q2_K":{"mode":"chat","base_model":"granite-code:20b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q3_K_L":{"mode":"chat","base_model":"granite-code:20b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q3_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q3_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q4_0":{"mode":"chat","base_model":"granite-code:20b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q4_1":{"mode":"chat","base_model":"granite-code:20b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q4_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q4_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q5_0":{"mode":"chat","base_model":"granite-code:20b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q5_1":{"mode":"chat","base_model":"granite-code:20b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q5_K_M":{"mode":"chat","base_model":"granite-code:20b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q5_K_S":{"mode":"chat","base_model":"granite-code:20b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q6_K":{"mode":"chat","base_model":"granite-code:20b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:20b-instruct-q8_0":{"mode":"chat","base_model":"granite-code:20b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base":{"mode":"chat","base_model":"granite-code:34b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q2_K":{"mode":"chat","base_model":"granite-code:34b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q3_K_L":{"mode":"chat","base_model":"granite-code:34b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q3_K_M":{"mode":"chat","base_model":"granite-code:34b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q3_K_S":{"mode":"chat","base_model":"granite-code:34b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q4_0":{"mode":"chat","base_model":"granite-code:34b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q4_1":{"mode":"chat","base_model":"granite-code:34b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q4_K_M":{"mode":"chat","base_model":"granite-code:34b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q4_K_S":{"mode":"chat","base_model":"granite-code:34b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q5_0":{"mode":"chat","base_model":"granite-code:34b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q5_1":{"mode":"chat","base_model":"granite-code:34b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q5_K_M":{"mode":"chat","base_model":"granite-code:34b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q5_K_S":{"mode":"chat","base_model":"granite-code:34b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q6_K":{"mode":"chat","base_model":"granite-code:34b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-base-q8_0":{"mode":"chat","base_model":"granite-code:34b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct":{"mode":"chat","base_model":"granite-code:34b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q2_K":{"mode":"chat","base_model":"granite-code:34b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q3_K_L":{"mode":"chat","base_model":"granite-code:34b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q3_K_M":{"mode":"chat","base_model":"granite-code:34b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q3_K_S":{"mode":"chat","base_model":"granite-code:34b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q4_0":{"mode":"chat","base_model":"granite-code:34b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q4_1":{"mode":"chat","base_model":"granite-code:34b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q4_K_M":{"mode":"chat","base_model":"granite-code:34b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q4_K_S":{"mode":"chat","base_model":"granite-code:34b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q5_0":{"mode":"chat","base_model":"granite-code:34b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q5_1":{"mode":"chat","base_model":"granite-code:34b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q5_K_M":{"mode":"chat","base_model":"granite-code:34b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q5_K_S":{"mode":"chat","base_model":"granite-code:34b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q6_K":{"mode":"chat","base_model":"granite-code:34b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:34b-instruct-q8_0":{"mode":"chat","base_model":"granite-code:34b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base":{"mode":"chat","base_model":"granite-code:3b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-fp16":{"mode":"chat","base_model":"granite-code:3b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q2_K":{"mode":"chat","base_model":"granite-code:3b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q3_K_L":{"mode":"chat","base_model":"granite-code:3b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q3_K_M":{"mode":"chat","base_model":"granite-code:3b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q3_K_S":{"mode":"chat","base_model":"granite-code:3b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q4_0":{"mode":"chat","base_model":"granite-code:3b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q4_1":{"mode":"chat","base_model":"granite-code:3b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q4_K_M":{"mode":"chat","base_model":"granite-code:3b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q4_K_S":{"mode":"chat","base_model":"granite-code:3b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q5_0":{"mode":"chat","base_model":"granite-code:3b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q5_1":{"mode":"chat","base_model":"granite-code:3b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q5_K_M":{"mode":"chat","base_model":"granite-code:3b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q5_K_S":{"mode":"chat","base_model":"granite-code:3b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q6_K":{"mode":"chat","base_model":"granite-code:3b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-base-q8_0":{"mode":"chat","base_model":"granite-code:3b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct":{"mode":"chat","base_model":"granite-code:3b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-fp16":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q2_K":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q3_K_L":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q3_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q3_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q4_0":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q4_1":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q4_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q4_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q5_0":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q5_1":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q5_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q5_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q6_K":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-128k-q8_0":{"mode":"chat","base_model":"granite-code:3b-instruct-128k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-fp16":{"mode":"chat","base_model":"granite-code:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q2_K":{"mode":"chat","base_model":"granite-code:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q3_K_L":{"mode":"chat","base_model":"granite-code:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q3_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q3_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q4_0":{"mode":"chat","base_model":"granite-code:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q4_1":{"mode":"chat","base_model":"granite-code:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q4_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q4_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q5_0":{"mode":"chat","base_model":"granite-code:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q5_1":{"mode":"chat","base_model":"granite-code:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q5_K_M":{"mode":"chat","base_model":"granite-code:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q5_K_S":{"mode":"chat","base_model":"granite-code:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q6_K":{"mode":"chat","base_model":"granite-code:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:3b-instruct-q8_0":{"mode":"chat","base_model":"granite-code:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base":{"mode":"chat","base_model":"granite-code:8b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-fp16":{"mode":"chat","base_model":"granite-code:8b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q2_K":{"mode":"chat","base_model":"granite-code:8b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q3_K_L":{"mode":"chat","base_model":"granite-code:8b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q3_K_M":{"mode":"chat","base_model":"granite-code:8b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q3_K_S":{"mode":"chat","base_model":"granite-code:8b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q4_0":{"mode":"chat","base_model":"granite-code:8b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q4_1":{"mode":"chat","base_model":"granite-code:8b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q4_K_M":{"mode":"chat","base_model":"granite-code:8b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q4_K_S":{"mode":"chat","base_model":"granite-code:8b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q5_0":{"mode":"chat","base_model":"granite-code:8b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q5_1":{"mode":"chat","base_model":"granite-code:8b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q5_K_M":{"mode":"chat","base_model":"granite-code:8b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q5_K_S":{"mode":"chat","base_model":"granite-code:8b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q6_K":{"mode":"chat","base_model":"granite-code:8b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-base-q8_0":{"mode":"chat","base_model":"granite-code:8b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct":{"mode":"chat","base_model":"granite-code:8b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-128k-q4_0":{"mode":"chat","base_model":"granite-code:8b-instruct-128k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-128k-q4_1":{"mode":"chat","base_model":"granite-code:8b-instruct-128k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-fp16":{"mode":"chat","base_model":"granite-code:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q2_K":{"mode":"chat","base_model":"granite-code:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q3_K_L":{"mode":"chat","base_model":"granite-code:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q3_K_M":{"mode":"chat","base_model":"granite-code:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q3_K_S":{"mode":"chat","base_model":"granite-code:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q4_0":{"mode":"chat","base_model":"granite-code:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q4_1":{"mode":"chat","base_model":"granite-code:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q4_K_M":{"mode":"chat","base_model":"granite-code:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q4_K_S":{"mode":"chat","base_model":"granite-code:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q5_0":{"mode":"chat","base_model":"granite-code:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q5_1":{"mode":"chat","base_model":"granite-code:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q5_K_M":{"mode":"chat","base_model":"granite-code:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q5_K_S":{"mode":"chat","base_model":"granite-code:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q6_K":{"mode":"chat","base_model":"granite-code:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-code:8b-instruct-q8_0":{"mode":"chat","base_model":"granite-code:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b":{"mode":"chat","base_model":"starcoder:1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b":{"mode":"chat","base_model":"starcoder:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b":{"mode":"chat","base_model":"starcoder:15b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base":{"mode":"chat","base_model":"starcoder:15b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-fp16":{"mode":"chat","base_model":"starcoder:15b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q2_K":{"mode":"chat","base_model":"starcoder:15b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q3_K_L":{"mode":"chat","base_model":"starcoder:15b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q3_K_M":{"mode":"chat","base_model":"starcoder:15b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q3_K_S":{"mode":"chat","base_model":"starcoder:15b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q4_0":{"mode":"chat","base_model":"starcoder:15b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q4_1":{"mode":"chat","base_model":"starcoder:15b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q4_K_M":{"mode":"chat","base_model":"starcoder:15b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q4_K_S":{"mode":"chat","base_model":"starcoder:15b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q5_0":{"mode":"chat","base_model":"starcoder:15b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q5_1":{"mode":"chat","base_model":"starcoder:15b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q5_K_M":{"mode":"chat","base_model":"starcoder:15b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q5_K_S":{"mode":"chat","base_model":"starcoder:15b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q6_K":{"mode":"chat","base_model":"starcoder:15b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-base-q8_0":{"mode":"chat","base_model":"starcoder:15b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-fp16":{"mode":"chat","base_model":"starcoder:15b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus":{"mode":"chat","base_model":"starcoder:15b-plus","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-fp16":{"mode":"chat","base_model":"starcoder:15b-plus-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q2_K":{"mode":"chat","base_model":"starcoder:15b-plus-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q3_K_L":{"mode":"chat","base_model":"starcoder:15b-plus-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q3_K_M":{"mode":"chat","base_model":"starcoder:15b-plus-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q3_K_S":{"mode":"chat","base_model":"starcoder:15b-plus-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q4_0":{"mode":"chat","base_model":"starcoder:15b-plus-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q4_1":{"mode":"chat","base_model":"starcoder:15b-plus-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q4_K_M":{"mode":"chat","base_model":"starcoder:15b-plus-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q4_K_S":{"mode":"chat","base_model":"starcoder:15b-plus-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q5_0":{"mode":"chat","base_model":"starcoder:15b-plus-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q5_1":{"mode":"chat","base_model":"starcoder:15b-plus-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q5_K_M":{"mode":"chat","base_model":"starcoder:15b-plus-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q5_K_S":{"mode":"chat","base_model":"starcoder:15b-plus-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q6_K":{"mode":"chat","base_model":"starcoder:15b-plus-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-plus-q8_0":{"mode":"chat","base_model":"starcoder:15b-plus-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q2_K":{"mode":"chat","base_model":"starcoder:15b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q3_K_L":{"mode":"chat","base_model":"starcoder:15b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q3_K_M":{"mode":"chat","base_model":"starcoder:15b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q3_K_S":{"mode":"chat","base_model":"starcoder:15b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q4_0":{"mode":"chat","base_model":"starcoder:15b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q4_1":{"mode":"chat","base_model":"starcoder:15b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q4_K_M":{"mode":"chat","base_model":"starcoder:15b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q4_K_S":{"mode":"chat","base_model":"starcoder:15b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q5_0":{"mode":"chat","base_model":"starcoder:15b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q5_1":{"mode":"chat","base_model":"starcoder:15b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q5_K_M":{"mode":"chat","base_model":"starcoder:15b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q5_K_S":{"mode":"chat","base_model":"starcoder:15b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q6_K":{"mode":"chat","base_model":"starcoder:15b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:15b-q8_0":{"mode":"chat","base_model":"starcoder:15b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base":{"mode":"chat","base_model":"starcoder:1b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-fp16":{"mode":"chat","base_model":"starcoder:1b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q2_K":{"mode":"chat","base_model":"starcoder:1b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q3_K_L":{"mode":"chat","base_model":"starcoder:1b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q3_K_M":{"mode":"chat","base_model":"starcoder:1b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q3_K_S":{"mode":"chat","base_model":"starcoder:1b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q4_0":{"mode":"chat","base_model":"starcoder:1b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q4_1":{"mode":"chat","base_model":"starcoder:1b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q4_K_M":{"mode":"chat","base_model":"starcoder:1b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q4_K_S":{"mode":"chat","base_model":"starcoder:1b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q5_0":{"mode":"chat","base_model":"starcoder:1b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q5_1":{"mode":"chat","base_model":"starcoder:1b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q5_K_M":{"mode":"chat","base_model":"starcoder:1b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q5_K_S":{"mode":"chat","base_model":"starcoder:1b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q6_K":{"mode":"chat","base_model":"starcoder:1b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:1b-base-q8_0":{"mode":"chat","base_model":"starcoder:1b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base":{"mode":"chat","base_model":"starcoder:3b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-fp16":{"mode":"chat","base_model":"starcoder:3b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q2_K":{"mode":"chat","base_model":"starcoder:3b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q3_K_L":{"mode":"chat","base_model":"starcoder:3b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q3_K_M":{"mode":"chat","base_model":"starcoder:3b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q3_K_S":{"mode":"chat","base_model":"starcoder:3b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q4_0":{"mode":"chat","base_model":"starcoder:3b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q4_1":{"mode":"chat","base_model":"starcoder:3b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q4_K_M":{"mode":"chat","base_model":"starcoder:3b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q4_K_S":{"mode":"chat","base_model":"starcoder:3b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q5_0":{"mode":"chat","base_model":"starcoder:3b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q5_1":{"mode":"chat","base_model":"starcoder:3b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q5_K_M":{"mode":"chat","base_model":"starcoder:3b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q5_K_S":{"mode":"chat","base_model":"starcoder:3b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q6_K":{"mode":"chat","base_model":"starcoder:3b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:3b-base-q8_0":{"mode":"chat","base_model":"starcoder:3b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base":{"mode":"chat","base_model":"starcoder:7b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-fp16":{"mode":"chat","base_model":"starcoder:7b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q2_K":{"mode":"chat","base_model":"starcoder:7b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q3_K_L":{"mode":"chat","base_model":"starcoder:7b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q3_K_M":{"mode":"chat","base_model":"starcoder:7b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q3_K_S":{"mode":"chat","base_model":"starcoder:7b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q4_0":{"mode":"chat","base_model":"starcoder:7b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q4_1":{"mode":"chat","base_model":"starcoder:7b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q4_K_M":{"mode":"chat","base_model":"starcoder:7b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q4_K_S":{"mode":"chat","base_model":"starcoder:7b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q5_0":{"mode":"chat","base_model":"starcoder:7b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q5_1":{"mode":"chat","base_model":"starcoder:7b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q5_K_M":{"mode":"chat","base_model":"starcoder:7b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q5_K_S":{"mode":"chat","base_model":"starcoder:7b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q6_K":{"mode":"chat","base_model":"starcoder:7b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starcoder:7b-base-q8_0":{"mode":"chat","base_model":"starcoder:7b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m":{"mode":"chat","base_model":"smollm:135m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m":{"mode":"chat","base_model":"smollm:360m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b":{"mode":"chat","base_model":"smollm:1.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-fp16":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q2_K":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q4_0":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q4_1":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q5_0":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q5_1":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q6_K":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-base-v0.2-q8_0":{"mode":"chat","base_model":"smollm:1.7b-base-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-fp16":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q2_K":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q4_0":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q4_1":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q5_0":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q5_1":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q6_K":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:1.7b-instruct-v0.2-q8_0":{"mode":"chat","base_model":"smollm:1.7b-instruct-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-fp16":{"mode":"chat","base_model":"smollm:135m-base-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q2_K":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q4_0":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q4_1":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q5_0":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q5_1":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q6_K":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-base-v0.2-q8_0":{"mode":"chat","base_model":"smollm:135m-base-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-fp16":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q2_K":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q4_0":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q4_1":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q5_0":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q5_1":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q6_K":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:135m-instruct-v0.2-q8_0":{"mode":"chat","base_model":"smollm:135m-instruct-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-fp16":{"mode":"chat","base_model":"smollm:360m-base-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q2_K":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q4_0":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q4_1":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q5_0":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q5_1":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q6_K":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-base-v0.2-q8_0":{"mode":"chat","base_model":"smollm:360m-base-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-fp16":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q2_K":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q3_K_L":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q3_K_M":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q3_K_S":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q4_0":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q4_1":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q4_K_M":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q4_K_S":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q5_0":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q5_1":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q5_K_M":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q5_K_S":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q6_K":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smollm:360m-instruct-v0.2-q8_0":{"mode":"chat","base_model":"smollm:360m-instruct-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-fp16":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q2_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q3_K_L":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q3_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q3_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q4_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q4_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q4_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q4_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q5_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q5_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q5_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q5_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q6_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:13b-q8_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-fp16":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q2_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q3_K_L":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q3_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q3_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q4_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q4_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q4_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q4_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q5_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q5_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q5_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q5_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q6_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:30b-q8_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:30b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-fp16":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q2_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q3_K_L":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q3_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q3_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q4_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q4_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q4_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q4_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q5_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q5_1":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q5_K_M":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q5_K_S":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q6_K":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna-uncensored:7b-q8_0":{"mode":"chat","base_model":"wizard-vicuna-uncensored:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b":{"mode":"chat","base_model":"vicuna:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b":{"mode":"chat","base_model":"vicuna:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b":{"mode":"chat","base_model":"vicuna:33b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-16k":{"mode":"chat","base_model":"vicuna:13b-16k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-fp16":{"mode":"chat","base_model":"vicuna:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q2_K":{"mode":"chat","base_model":"vicuna:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q3_K_L":{"mode":"chat","base_model":"vicuna:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q3_K_M":{"mode":"chat","base_model":"vicuna:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q3_K_S":{"mode":"chat","base_model":"vicuna:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q4_0":{"mode":"chat","base_model":"vicuna:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q4_1":{"mode":"chat","base_model":"vicuna:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q4_K_M":{"mode":"chat","base_model":"vicuna:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q4_K_S":{"mode":"chat","base_model":"vicuna:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q5_0":{"mode":"chat","base_model":"vicuna:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q5_1":{"mode":"chat","base_model":"vicuna:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q5_K_M":{"mode":"chat","base_model":"vicuna:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q5_K_S":{"mode":"chat","base_model":"vicuna:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q6_K":{"mode":"chat","base_model":"vicuna:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-q8_0":{"mode":"chat","base_model":"vicuna:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-fp16":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q2_K":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q3_K_L":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q3_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q3_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q4_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q4_1":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q4_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q4_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q5_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q5_1":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q5_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q5_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q6_K":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-16k-q8_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-16k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-fp16":{"mode":"chat","base_model":"vicuna:13b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q2_K":{"mode":"chat","base_model":"vicuna:13b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q3_K_L":{"mode":"chat","base_model":"vicuna:13b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q3_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q3_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q4_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q4_1":{"mode":"chat","base_model":"vicuna:13b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q4_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q4_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q5_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q5_1":{"mode":"chat","base_model":"vicuna:13b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q5_K_M":{"mode":"chat","base_model":"vicuna:13b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q5_K_S":{"mode":"chat","base_model":"vicuna:13b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q6_K":{"mode":"chat","base_model":"vicuna:13b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:13b-v1.5-q8_0":{"mode":"chat","base_model":"vicuna:13b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-fp16":{"mode":"chat","base_model":"vicuna:33b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q2_K":{"mode":"chat","base_model":"vicuna:33b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q3_K_L":{"mode":"chat","base_model":"vicuna:33b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q3_K_M":{"mode":"chat","base_model":"vicuna:33b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q3_K_S":{"mode":"chat","base_model":"vicuna:33b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q4_0":{"mode":"chat","base_model":"vicuna:33b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q4_1":{"mode":"chat","base_model":"vicuna:33b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q4_K_M":{"mode":"chat","base_model":"vicuna:33b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q4_K_S":{"mode":"chat","base_model":"vicuna:33b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q5_0":{"mode":"chat","base_model":"vicuna:33b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q5_1":{"mode":"chat","base_model":"vicuna:33b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q5_K_M":{"mode":"chat","base_model":"vicuna:33b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q5_K_S":{"mode":"chat","base_model":"vicuna:33b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q6_K":{"mode":"chat","base_model":"vicuna:33b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:33b-q8_0":{"mode":"chat","base_model":"vicuna:33b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-16k":{"mode":"chat","base_model":"vicuna:7b-16k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-fp16":{"mode":"chat","base_model":"vicuna:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q2_K":{"mode":"chat","base_model":"vicuna:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q3_K_L":{"mode":"chat","base_model":"vicuna:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q3_K_M":{"mode":"chat","base_model":"vicuna:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q3_K_S":{"mode":"chat","base_model":"vicuna:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q4_0":{"mode":"chat","base_model":"vicuna:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q4_1":{"mode":"chat","base_model":"vicuna:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q4_K_M":{"mode":"chat","base_model":"vicuna:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q4_K_S":{"mode":"chat","base_model":"vicuna:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q5_0":{"mode":"chat","base_model":"vicuna:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q5_1":{"mode":"chat","base_model":"vicuna:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q5_K_M":{"mode":"chat","base_model":"vicuna:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q5_K_S":{"mode":"chat","base_model":"vicuna:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q6_K":{"mode":"chat","base_model":"vicuna:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-q8_0":{"mode":"chat","base_model":"vicuna:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-fp16":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q2_K":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q3_K_L":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q3_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q3_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q4_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q4_1":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q4_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q4_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q5_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q5_1":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q5_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q5_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q6_K":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-16k-q8_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-16k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-fp16":{"mode":"chat","base_model":"vicuna:7b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q2_K":{"mode":"chat","base_model":"vicuna:7b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q3_K_L":{"mode":"chat","base_model":"vicuna:7b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q3_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q3_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q4_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q4_1":{"mode":"chat","base_model":"vicuna:7b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q4_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q4_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q5_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q5_1":{"mode":"chat","base_model":"vicuna:7b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q5_K_M":{"mode":"chat","base_model":"vicuna:7b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q5_K_S":{"mode":"chat","base_model":"vicuna:7b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q6_K":{"mode":"chat","base_model":"vicuna:7b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"vicuna:7b-v1.5-q8_0":{"mode":"chat","base_model":"vicuna:7b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b":{"mode":"chat","base_model":"mistral-openorca:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-fp16":{"mode":"chat","base_model":"mistral-openorca:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q2_K":{"mode":"chat","base_model":"mistral-openorca:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q3_K_L":{"mode":"chat","base_model":"mistral-openorca:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q3_K_M":{"mode":"chat","base_model":"mistral-openorca:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q3_K_S":{"mode":"chat","base_model":"mistral-openorca:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q4_0":{"mode":"chat","base_model":"mistral-openorca:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q4_1":{"mode":"chat","base_model":"mistral-openorca:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q4_K_M":{"mode":"chat","base_model":"mistral-openorca:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q4_K_S":{"mode":"chat","base_model":"mistral-openorca:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q5_0":{"mode":"chat","base_model":"mistral-openorca:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q5_1":{"mode":"chat","base_model":"mistral-openorca:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q5_K_M":{"mode":"chat","base_model":"mistral-openorca:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q5_K_S":{"mode":"chat","base_model":"mistral-openorca:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q6_K":{"mode":"chat","base_model":"mistral-openorca:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-openorca:7b-q8_0":{"mode":"chat","base_model":"mistral-openorca:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwq:32b-preview-fp16":{"mode":"chat","base_model":"qwq:32b-preview-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwq:32b-preview-q4_K_M":{"mode":"chat","base_model":"qwq:32b-preview-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwq:32b-preview-q8_0":{"mode":"chat","base_model":"qwq:32b-preview-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b":{"mode":"chat","base_model":"llama2-chinese:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b":{"mode":"chat","base_model":"llama2-chinese:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat":{"mode":"chat","base_model":"llama2-chinese:13b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-fp16":{"mode":"chat","base_model":"llama2-chinese:13b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q2_K":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q3_K_L":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q3_K_M":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q3_K_S":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q4_0":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q4_1":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q4_K_M":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q4_K_S":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q5_0":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q5_1":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q5_K_M":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q5_K_S":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q6_K":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:13b-chat-q8_0":{"mode":"chat","base_model":"llama2-chinese:13b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat":{"mode":"chat","base_model":"llama2-chinese:7b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-fp16":{"mode":"chat","base_model":"llama2-chinese:7b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q2_K":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q3_K_L":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q3_K_M":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q3_K_S":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q4_0":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q4_1":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q4_K_M":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q4_K_S":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q5_0":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q5_1":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q5_K_M":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q5_K_S":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q6_K":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama2-chinese:7b-chat-q8_0":{"mode":"chat","base_model":"llama2-chinese:7b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b":{"mode":"chat","base_model":"openchat:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5":{"mode":"chat","base_model":"openchat:7b-v3.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106":{"mode":"chat","base_model":"openchat:7b-v3.5-0106","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-fp16":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q2_K":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q3_K_L":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q3_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q3_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q4_0":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q4_1":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q4_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q4_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q5_0":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q5_1":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q5_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q5_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q6_K":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-0106-q8_0":{"mode":"chat","base_model":"openchat:7b-v3.5-0106-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210":{"mode":"chat","base_model":"openchat:7b-v3.5-1210","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-fp16":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q2_K":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q3_K_L":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q3_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q3_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q4_0":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q4_1":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q4_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q4_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q5_0":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q5_1":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q5_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q5_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q6_K":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-1210-q8_0":{"mode":"chat","base_model":"openchat:7b-v3.5-1210-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-fp16":{"mode":"chat","base_model":"openchat:7b-v3.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q2_K":{"mode":"chat","base_model":"openchat:7b-v3.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q3_K_L":{"mode":"chat","base_model":"openchat:7b-v3.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q3_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q3_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q4_0":{"mode":"chat","base_model":"openchat:7b-v3.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q4_1":{"mode":"chat","base_model":"openchat:7b-v3.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q4_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q4_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q5_0":{"mode":"chat","base_model":"openchat:7b-v3.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q5_1":{"mode":"chat","base_model":"openchat:7b-v3.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q5_K_M":{"mode":"chat","base_model":"openchat:7b-v3.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q5_K_S":{"mode":"chat","base_model":"openchat:7b-v3.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q6_K":{"mode":"chat","base_model":"openchat:7b-v3.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openchat:7b-v3.5-q8_0":{"mode":"chat","base_model":"openchat:7b-v3.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b":{"mode":"chat","base_model":"codegeex4:9b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-fp16":{"mode":"chat","base_model":"codegeex4:9b-all-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q2_K":{"mode":"chat","base_model":"codegeex4:9b-all-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q3_K_L":{"mode":"chat","base_model":"codegeex4:9b-all-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q3_K_M":{"mode":"chat","base_model":"codegeex4:9b-all-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q3_K_S":{"mode":"chat","base_model":"codegeex4:9b-all-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q4_0":{"mode":"chat","base_model":"codegeex4:9b-all-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q4_1":{"mode":"chat","base_model":"codegeex4:9b-all-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q4_K_M":{"mode":"chat","base_model":"codegeex4:9b-all-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q4_K_S":{"mode":"chat","base_model":"codegeex4:9b-all-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q5_0":{"mode":"chat","base_model":"codegeex4:9b-all-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q5_1":{"mode":"chat","base_model":"codegeex4:9b-all-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q5_K_M":{"mode":"chat","base_model":"codegeex4:9b-all-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q5_K_S":{"mode":"chat","base_model":"codegeex4:9b-all-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q6_K":{"mode":"chat","base_model":"codegeex4:9b-all-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codegeex4:9b-all-q8_0":{"mode":"chat","base_model":"codegeex4:9b-all-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b":{"mode":"chat","base_model":"aya:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b":{"mode":"chat","base_model":"aya:35b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23":{"mode":"chat","base_model":"aya:35b-23","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q2_K":{"mode":"chat","base_model":"aya:35b-23-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q3_K_L":{"mode":"chat","base_model":"aya:35b-23-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q3_K_M":{"mode":"chat","base_model":"aya:35b-23-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q3_K_S":{"mode":"chat","base_model":"aya:35b-23-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q4_0":{"mode":"chat","base_model":"aya:35b-23-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q4_1":{"mode":"chat","base_model":"aya:35b-23-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q4_K_M":{"mode":"chat","base_model":"aya:35b-23-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q4_K_S":{"mode":"chat","base_model":"aya:35b-23-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q5_0":{"mode":"chat","base_model":"aya:35b-23-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q5_1":{"mode":"chat","base_model":"aya:35b-23-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q5_K_M":{"mode":"chat","base_model":"aya:35b-23-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q5_K_S":{"mode":"chat","base_model":"aya:35b-23-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q6_K":{"mode":"chat","base_model":"aya:35b-23-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:35b-23-q8_0":{"mode":"chat","base_model":"aya:35b-23-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23":{"mode":"chat","base_model":"aya:8b-23","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q2_K":{"mode":"chat","base_model":"aya:8b-23-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q3_K_L":{"mode":"chat","base_model":"aya:8b-23-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q3_K_M":{"mode":"chat","base_model":"aya:8b-23-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q3_K_S":{"mode":"chat","base_model":"aya:8b-23-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q4_0":{"mode":"chat","base_model":"aya:8b-23-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q4_1":{"mode":"chat","base_model":"aya:8b-23-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q4_K_M":{"mode":"chat","base_model":"aya:8b-23-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q4_K_S":{"mode":"chat","base_model":"aya:8b-23-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q5_0":{"mode":"chat","base_model":"aya:8b-23-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q5_1":{"mode":"chat","base_model":"aya:8b-23-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q5_K_M":{"mode":"chat","base_model":"aya:8b-23-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q5_K_S":{"mode":"chat","base_model":"aya:8b-23-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q6_K":{"mode":"chat","base_model":"aya:8b-23-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya:8b-23-q8_0":{"mode":"chat","base_model":"aya:8b-23-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b":{"mode":"chat","base_model":"codeqwen:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat":{"mode":"chat","base_model":"codeqwen:7b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-fp16":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q2_K":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q3_K_L":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q3_K_M":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q3_K_S":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q4_0":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q4_1":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q4_K_M":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q4_K_S":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q5_0":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q5_1":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q5_K_M":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q5_K_S":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q6_K":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-chat-v1.5-q8_0":{"mode":"chat","base_model":"codeqwen:7b-chat-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code":{"mode":"chat","base_model":"codeqwen:7b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-fp16":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-q4_0":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-q4_1":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-q5_0":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-q5_1":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:7b-code-v1.5-q8_0":{"mode":"chat","base_model":"codeqwen:7b-code-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:chat":{"mode":"chat","base_model":"codeqwen:chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:code":{"mode":"chat","base_model":"codeqwen:code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"codeqwen:v1.5":{"mode":"chat","base_model":"codeqwen:v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:v1.5-chat":{"mode":"chat","base_model":"codeqwen:v1.5-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeqwen:v1.5-code":{"mode":"chat","base_model":"codeqwen:v1.5-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b":{"mode":"chat","base_model":"deepseek-llm:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b":{"mode":"chat","base_model":"deepseek-llm:67b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base":{"mode":"chat","base_model":"deepseek-llm:67b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-fp16":{"mode":"chat","base_model":"deepseek-llm:67b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q2_K":{"mode":"chat","base_model":"deepseek-llm:67b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-llm:67b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-llm:67b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q4_0":{"mode":"chat","base_model":"deepseek-llm:67b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q4_1":{"mode":"chat","base_model":"deepseek-llm:67b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-llm:67b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q5_0":{"mode":"chat","base_model":"deepseek-llm:67b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q5_1":{"mode":"chat","base_model":"deepseek-llm:67b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-llm:67b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q6_K":{"mode":"chat","base_model":"deepseek-llm:67b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-base-q8_0":{"mode":"chat","base_model":"deepseek-llm:67b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat":{"mode":"chat","base_model":"deepseek-llm:67b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-fp16":{"mode":"chat","base_model":"deepseek-llm:67b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q2_K":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q3_K_L":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q3_K_M":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q3_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q4_0":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q4_1":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q4_K_M":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q4_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q5_0":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q5_1":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:67b-chat-q5_K_S":{"mode":"chat","base_model":"deepseek-llm:67b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base":{"mode":"chat","base_model":"deepseek-llm:7b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-fp16":{"mode":"chat","base_model":"deepseek-llm:7b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q2_K":{"mode":"chat","base_model":"deepseek-llm:7b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q3_K_L":{"mode":"chat","base_model":"deepseek-llm:7b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q3_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q3_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q4_0":{"mode":"chat","base_model":"deepseek-llm:7b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q4_1":{"mode":"chat","base_model":"deepseek-llm:7b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q4_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q4_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q5_0":{"mode":"chat","base_model":"deepseek-llm:7b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q5_1":{"mode":"chat","base_model":"deepseek-llm:7b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q5_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q5_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q6_K":{"mode":"chat","base_model":"deepseek-llm:7b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-base-q8_0":{"mode":"chat","base_model":"deepseek-llm:7b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat":{"mode":"chat","base_model":"deepseek-llm:7b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-fp16":{"mode":"chat","base_model":"deepseek-llm:7b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q2_K":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q3_K_L":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q3_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q3_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q4_0":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q4_1":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q4_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q4_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q5_0":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q5_1":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q5_K_M":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q5_K_S":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q6_K":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-llm:7b-chat-q8_0":{"mode":"chat","base_model":"deepseek-llm:7b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b":{"mode":"chat","base_model":"deepseek-v2:16b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b":{"mode":"chat","base_model":"deepseek-v2:236b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-fp16":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q2_K":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q3_K_L":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q3_K_M":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q3_K_S":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q4_0":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q4_1":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q4_K_M":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q4_K_S":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q5_0":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q5_1":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q5_K_M":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q5_K_S":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q6_K":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:16b-lite-chat-q8_0":{"mode":"chat","base_model":"deepseek-v2:16b-lite-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-fp16":{"mode":"chat","base_model":"deepseek-v2:236b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q2_K":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q3_K_L":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q3_K_M":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q3_K_S":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q4_0":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q4_1":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q4_K_M":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q4_K_S":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q5_0":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q5_1":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q5_K_M":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q5_K_S":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q6_K":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:236b-chat-q8_0":{"mode":"chat","base_model":"deepseek-v2:236b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2:lite":{"mode":"chat","base_model":"deepseek-v2:lite","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b":{"mode":"chat","base_model":"mistral-large:123b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-fp16":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q2_K":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q3_K_L":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q3_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q3_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q4_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q4_1":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q4_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q4_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q5_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q5_1":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q5_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q5_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q6_K":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2407-q8_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2407-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-fp16":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q2_K":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q3_K_L":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q3_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q3_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q4_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q4_1":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q4_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q4_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q5_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q5_1":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q5_K_M":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q5_K_S":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q6_K":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral-large:123b-instruct-2411-q8_0":{"mode":"chat","base_model":"mistral-large:123b-instruct-2411-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b":{"mode":"chat","base_model":"glm4:9b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-fp16":{"mode":"chat","base_model":"glm4:9b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q2_K":{"mode":"chat","base_model":"glm4:9b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q3_K_L":{"mode":"chat","base_model":"glm4:9b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q3_K_M":{"mode":"chat","base_model":"glm4:9b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q3_K_S":{"mode":"chat","base_model":"glm4:9b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q4_0":{"mode":"chat","base_model":"glm4:9b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q4_1":{"mode":"chat","base_model":"glm4:9b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q4_K_M":{"mode":"chat","base_model":"glm4:9b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q4_K_S":{"mode":"chat","base_model":"glm4:9b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q5_0":{"mode":"chat","base_model":"glm4:9b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q5_1":{"mode":"chat","base_model":"glm4:9b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q5_K_M":{"mode":"chat","base_model":"glm4:9b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q5_K_S":{"mode":"chat","base_model":"glm4:9b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q6_K":{"mode":"chat","base_model":"glm4:9b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-chat-q8_0":{"mode":"chat","base_model":"glm4:9b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-fp16":{"mode":"chat","base_model":"glm4:9b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q2_K":{"mode":"chat","base_model":"glm4:9b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q3_K_L":{"mode":"chat","base_model":"glm4:9b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q3_K_M":{"mode":"chat","base_model":"glm4:9b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q3_K_S":{"mode":"chat","base_model":"glm4:9b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q4_0":{"mode":"chat","base_model":"glm4:9b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q4_1":{"mode":"chat","base_model":"glm4:9b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q4_K_M":{"mode":"chat","base_model":"glm4:9b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q4_K_S":{"mode":"chat","base_model":"glm4:9b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q5_0":{"mode":"chat","base_model":"glm4:9b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q5_1":{"mode":"chat","base_model":"glm4:9b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q5_K_M":{"mode":"chat","base_model":"glm4:9b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q5_K_S":{"mode":"chat","base_model":"glm4:9b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q6_K":{"mode":"chat","base_model":"glm4:9b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"glm4:9b-text-q8_0":{"mode":"chat","base_model":"glm4:9b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b":{"mode":"chat","base_model":"nous-hermes2:10.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b":{"mode":"chat","base_model":"nous-hermes2:34b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-fp16":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q2_K":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q3_K_L":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q3_K_M":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q3_K_S":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q4_0":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q4_1":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q4_K_M":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q4_K_S":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q5_0":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q5_1":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q5_K_M":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q5_K_S":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q6_K":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:10.7b-solar-q8_0":{"mode":"chat","base_model":"nous-hermes2:10.7b-solar-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-fp16":{"mode":"chat","base_model":"nous-hermes2:34b-yi-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q2_K":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q3_K_L":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q3_K_M":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q3_K_S":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q4_0":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q4_1":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q4_K_M":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q4_K_S":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q5_0":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q5_1":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q5_K_M":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q5_K_S":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q6_K":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2:34b-yi-q8_0":{"mode":"chat","base_model":"nous-hermes2:34b-yi-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code":{"mode":"chat","base_model":"stable-code:3b-code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-fp16":{"mode":"chat","base_model":"stable-code:3b-code-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q2_K":{"mode":"chat","base_model":"stable-code:3b-code-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q3_K_L":{"mode":"chat","base_model":"stable-code:3b-code-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q3_K_M":{"mode":"chat","base_model":"stable-code:3b-code-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q3_K_S":{"mode":"chat","base_model":"stable-code:3b-code-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q4_0":{"mode":"chat","base_model":"stable-code:3b-code-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q4_1":{"mode":"chat","base_model":"stable-code:3b-code-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q4_K_M":{"mode":"chat","base_model":"stable-code:3b-code-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q4_K_S":{"mode":"chat","base_model":"stable-code:3b-code-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q5_0":{"mode":"chat","base_model":"stable-code:3b-code-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q5_1":{"mode":"chat","base_model":"stable-code:3b-code-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q5_K_M":{"mode":"chat","base_model":"stable-code:3b-code-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q5_K_S":{"mode":"chat","base_model":"stable-code:3b-code-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q6_K":{"mode":"chat","base_model":"stable-code:3b-code-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-code-q8_0":{"mode":"chat","base_model":"stable-code:3b-code-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct":{"mode":"chat","base_model":"stable-code:3b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-fp16":{"mode":"chat","base_model":"stable-code:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q2_K":{"mode":"chat","base_model":"stable-code:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q3_K_L":{"mode":"chat","base_model":"stable-code:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q3_K_M":{"mode":"chat","base_model":"stable-code:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q3_K_S":{"mode":"chat","base_model":"stable-code:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q4_0":{"mode":"chat","base_model":"stable-code:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q4_1":{"mode":"chat","base_model":"stable-code:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q4_K_M":{"mode":"chat","base_model":"stable-code:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q4_K_S":{"mode":"chat","base_model":"stable-code:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q5_0":{"mode":"chat","base_model":"stable-code:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q5_1":{"mode":"chat","base_model":"stable-code:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q5_K_M":{"mode":"chat","base_model":"stable-code:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q5_K_S":{"mode":"chat","base_model":"stable-code:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q6_K":{"mode":"chat","base_model":"stable-code:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:3b-instruct-q8_0":{"mode":"chat","base_model":"stable-code:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:code":{"mode":"chat","base_model":"stable-code:code","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-code:instruct":{"mode":"chat","base_model":"stable-code:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-fp16":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q2_K":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q3_K_L":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q3_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q3_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q4_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q4_1":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q4_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q4_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q5_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q5_1":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q5_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q5_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q6_K":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2-q8_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-fp16":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q2_K":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q3_K_L":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q3_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q3_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q4_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q4_1":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q4_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q4_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q5_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q5_1":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q5_K_M":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q5_K_S":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q6_K":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-mistral-v2.5-q8_0":{"mode":"chat","base_model":"openhermes:7b-mistral-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-v2":{"mode":"chat","base_model":"openhermes:7b-v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:7b-v2.5":{"mode":"chat","base_model":"openhermes:7b-v2.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:v2":{"mode":"chat","base_model":"openhermes:v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openhermes:v2.5":{"mode":"chat","base_model":"openhermes:v2.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b":{"mode":"chat","base_model":"qwen2-math:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b":{"mode":"chat","base_model":"qwen2-math:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b":{"mode":"chat","base_model":"qwen2-math:72b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-fp16":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q2_K":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q4_0":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q4_1":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q5_0":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q5_1":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q6_K":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:1.5b-instruct-q8_0":{"mode":"chat","base_model":"qwen2-math:1.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct":{"mode":"chat","base_model":"qwen2-math:72b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-fp16":{"mode":"chat","base_model":"qwen2-math:72b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q2_K":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q4_0":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q4_1":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q5_0":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q5_1":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q6_K":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:72b-instruct-q8_0":{"mode":"chat","base_model":"qwen2-math:72b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct":{"mode":"chat","base_model":"qwen2-math:7b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-fp16":{"mode":"chat","base_model":"qwen2-math:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q2_K":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q3_K_L":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q3_K_M":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q3_K_S":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q4_0":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q4_1":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q4_K_M":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q4_K_S":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q5_0":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q5_1":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q5_K_M":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q5_K_S":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q6_K":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"qwen2-math:7b-instruct-q8_0":{"mode":"chat","base_model":"qwen2-math:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b":{"mode":"chat","base_model":"tinydolphin:1.1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-fp16":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q2_K":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q3_K_L":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q3_K_M":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q3_K_S":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q4_0":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q4_1":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q4_K_M":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q4_K_S":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q5_0":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q5_1":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q5_K_M":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q5_K_S":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q6_K":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:1.1b-v2.8-q8_0":{"mode":"chat","base_model":"tinydolphin:1.1b-v2.8-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tinydolphin:v2.8":{"mode":"chat","base_model":"tinydolphin:v2.8","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b":{"mode":"chat","base_model":"command-r-plus:104b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-fp16":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q2_K":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q3_K_L":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q3_K_M":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q3_K_S":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q4_0":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q4_1":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q4_K_M":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q4_K_S":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q5_0":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q5_1":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q5_K_M":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q5_K_S":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q6_K":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-08-2024-q8_0":{"mode":"chat","base_model":"command-r-plus:104b-08-2024-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-fp16":{"mode":"chat","base_model":"command-r-plus:104b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-q2_K":{"mode":"chat","base_model":"command-r-plus:104b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-q4_0":{"mode":"chat","base_model":"command-r-plus:104b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r-plus:104b-q8_0":{"mode":"chat","base_model":"command-r-plus:104b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b":{"mode":"chat","base_model":"wizardcoder:33b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python":{"mode":"chat","base_model":"wizardcoder:13b-python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-fp16":{"mode":"chat","base_model":"wizardcoder:13b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q2_K":{"mode":"chat","base_model":"wizardcoder:13b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q3_K_L":{"mode":"chat","base_model":"wizardcoder:13b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q3_K_M":{"mode":"chat","base_model":"wizardcoder:13b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q3_K_S":{"mode":"chat","base_model":"wizardcoder:13b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q4_0":{"mode":"chat","base_model":"wizardcoder:13b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q4_1":{"mode":"chat","base_model":"wizardcoder:13b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q4_K_M":{"mode":"chat","base_model":"wizardcoder:13b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q4_K_S":{"mode":"chat","base_model":"wizardcoder:13b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q5_0":{"mode":"chat","base_model":"wizardcoder:13b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q5_1":{"mode":"chat","base_model":"wizardcoder:13b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q5_K_M":{"mode":"chat","base_model":"wizardcoder:13b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q5_K_S":{"mode":"chat","base_model":"wizardcoder:13b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q6_K":{"mode":"chat","base_model":"wizardcoder:13b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:13b-python-q8_0":{"mode":"chat","base_model":"wizardcoder:13b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1":{"mode":"chat","base_model":"wizardcoder:33b-v1.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-fp16":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q2_K":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q3_K_L":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q3_K_M":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q3_K_S":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q4_0":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q4_1":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q4_K_M":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q4_K_S":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q5_0":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q5_1":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q5_K_M":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q5_K_S":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q6_K":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:33b-v1.1-q8_0":{"mode":"chat","base_model":"wizardcoder:33b-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python":{"mode":"chat","base_model":"wizardcoder:34b-python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-fp16":{"mode":"chat","base_model":"wizardcoder:34b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q2_K":{"mode":"chat","base_model":"wizardcoder:34b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q3_K_L":{"mode":"chat","base_model":"wizardcoder:34b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q3_K_M":{"mode":"chat","base_model":"wizardcoder:34b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q3_K_S":{"mode":"chat","base_model":"wizardcoder:34b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q4_0":{"mode":"chat","base_model":"wizardcoder:34b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q4_1":{"mode":"chat","base_model":"wizardcoder:34b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q4_K_M":{"mode":"chat","base_model":"wizardcoder:34b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q4_K_S":{"mode":"chat","base_model":"wizardcoder:34b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q5_0":{"mode":"chat","base_model":"wizardcoder:34b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q5_1":{"mode":"chat","base_model":"wizardcoder:34b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q5_K_M":{"mode":"chat","base_model":"wizardcoder:34b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q5_K_S":{"mode":"chat","base_model":"wizardcoder:34b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q6_K":{"mode":"chat","base_model":"wizardcoder:34b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:34b-python-q8_0":{"mode":"chat","base_model":"wizardcoder:34b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python":{"mode":"chat","base_model":"wizardcoder:7b-python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-fp16":{"mode":"chat","base_model":"wizardcoder:7b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q2_K":{"mode":"chat","base_model":"wizardcoder:7b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q3_K_L":{"mode":"chat","base_model":"wizardcoder:7b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q3_K_M":{"mode":"chat","base_model":"wizardcoder:7b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q3_K_S":{"mode":"chat","base_model":"wizardcoder:7b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q4_0":{"mode":"chat","base_model":"wizardcoder:7b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q4_1":{"mode":"chat","base_model":"wizardcoder:7b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q4_K_M":{"mode":"chat","base_model":"wizardcoder:7b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q4_K_S":{"mode":"chat","base_model":"wizardcoder:7b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q5_0":{"mode":"chat","base_model":"wizardcoder:7b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q5_1":{"mode":"chat","base_model":"wizardcoder:7b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q5_K_M":{"mode":"chat","base_model":"wizardcoder:7b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q5_K_S":{"mode":"chat","base_model":"wizardcoder:7b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q6_K":{"mode":"chat","base_model":"wizardcoder:7b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:7b-python-q8_0":{"mode":"chat","base_model":"wizardcoder:7b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardcoder:python":{"mode":"chat","base_model":"wizardcoder:python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b":{"mode":"chat","base_model":"moondream:1.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-fp16":{"mode":"chat","base_model":"moondream:1.8b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q2_K":{"mode":"chat","base_model":"moondream:1.8b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q3_K_L":{"mode":"chat","base_model":"moondream:1.8b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q3_K_M":{"mode":"chat","base_model":"moondream:1.8b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q3_K_S":{"mode":"chat","base_model":"moondream:1.8b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q4_0":{"mode":"chat","base_model":"moondream:1.8b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q4_1":{"mode":"chat","base_model":"moondream:1.8b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q4_K_M":{"mode":"chat","base_model":"moondream:1.8b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q4_K_S":{"mode":"chat","base_model":"moondream:1.8b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q5_0":{"mode":"chat","base_model":"moondream:1.8b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q5_1":{"mode":"chat","base_model":"moondream:1.8b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q5_K_M":{"mode":"chat","base_model":"moondream:1.8b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q5_K_S":{"mode":"chat","base_model":"moondream:1.8b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q6_K":{"mode":"chat","base_model":"moondream:1.8b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:1.8b-v2-q8_0":{"mode":"chat","base_model":"moondream:1.8b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"moondream:v2":{"mode":"chat","base_model":"moondream:v2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b":{"mode":"chat","base_model":"bakllava:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-fp16":{"mode":"chat","base_model":"bakllava:7b-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q2_K":{"mode":"chat","base_model":"bakllava:7b-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q3_K_L":{"mode":"chat","base_model":"bakllava:7b-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q3_K_M":{"mode":"chat","base_model":"bakllava:7b-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q3_K_S":{"mode":"chat","base_model":"bakllava:7b-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q4_0":{"mode":"chat","base_model":"bakllava:7b-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q4_1":{"mode":"chat","base_model":"bakllava:7b-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q4_K_M":{"mode":"chat","base_model":"bakllava:7b-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q4_K_S":{"mode":"chat","base_model":"bakllava:7b-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q5_0":{"mode":"chat","base_model":"bakllava:7b-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q5_1":{"mode":"chat","base_model":"bakllava:7b-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q5_K_M":{"mode":"chat","base_model":"bakllava:7b-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q5_K_S":{"mode":"chat","base_model":"bakllava:7b-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q6_K":{"mode":"chat","base_model":"bakllava:7b-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bakllava:7b-v1-q8_0":{"mode":"chat","base_model":"bakllava:7b-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b":{"mode":"chat","base_model":"stablelm2:1.6b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b":{"mode":"chat","base_model":"stablelm2:12b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat":{"mode":"chat","base_model":"stablelm2:1.6b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-fp16":{"mode":"chat","base_model":"stablelm2:1.6b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q2_K":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q3_K_L":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q3_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q3_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q4_0":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q4_1":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q4_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q4_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q5_0":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q5_1":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q5_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q5_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q6_K":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-chat-q8_0":{"mode":"chat","base_model":"stablelm2:1.6b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-fp16":{"mode":"chat","base_model":"stablelm2:1.6b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q2_K":{"mode":"chat","base_model":"stablelm2:1.6b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q3_K_L":{"mode":"chat","base_model":"stablelm2:1.6b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q3_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q3_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q4_0":{"mode":"chat","base_model":"stablelm2:1.6b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q4_1":{"mode":"chat","base_model":"stablelm2:1.6b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q4_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q4_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q5_0":{"mode":"chat","base_model":"stablelm2:1.6b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q5_1":{"mode":"chat","base_model":"stablelm2:1.6b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q5_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q5_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q6_K":{"mode":"chat","base_model":"stablelm2:1.6b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-q8_0":{"mode":"chat","base_model":"stablelm2:1.6b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-fp16":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q2_K":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q3_K_L":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q3_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q3_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q4_0":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q4_1":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q4_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q4_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q5_0":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q5_1":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q5_K_M":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q5_K_S":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q6_K":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:1.6b-zephyr-q8_0":{"mode":"chat","base_model":"stablelm2:1.6b-zephyr-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat":{"mode":"chat","base_model":"stablelm2:12b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-fp16":{"mode":"chat","base_model":"stablelm2:12b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q2_K":{"mode":"chat","base_model":"stablelm2:12b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q3_K_L":{"mode":"chat","base_model":"stablelm2:12b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q3_K_M":{"mode":"chat","base_model":"stablelm2:12b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q3_K_S":{"mode":"chat","base_model":"stablelm2:12b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q4_0":{"mode":"chat","base_model":"stablelm2:12b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q4_1":{"mode":"chat","base_model":"stablelm2:12b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q4_K_M":{"mode":"chat","base_model":"stablelm2:12b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q4_K_S":{"mode":"chat","base_model":"stablelm2:12b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q5_0":{"mode":"chat","base_model":"stablelm2:12b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q5_1":{"mode":"chat","base_model":"stablelm2:12b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q5_K_M":{"mode":"chat","base_model":"stablelm2:12b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q5_K_S":{"mode":"chat","base_model":"stablelm2:12b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q6_K":{"mode":"chat","base_model":"stablelm2:12b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-chat-q8_0":{"mode":"chat","base_model":"stablelm2:12b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-fp16":{"mode":"chat","base_model":"stablelm2:12b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q2_K":{"mode":"chat","base_model":"stablelm2:12b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q3_K_L":{"mode":"chat","base_model":"stablelm2:12b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q3_K_M":{"mode":"chat","base_model":"stablelm2:12b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q3_K_S":{"mode":"chat","base_model":"stablelm2:12b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q4_0":{"mode":"chat","base_model":"stablelm2:12b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q4_1":{"mode":"chat","base_model":"stablelm2:12b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q4_K_M":{"mode":"chat","base_model":"stablelm2:12b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q4_K_S":{"mode":"chat","base_model":"stablelm2:12b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q5_0":{"mode":"chat","base_model":"stablelm2:12b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q5_1":{"mode":"chat","base_model":"stablelm2:12b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q5_K_M":{"mode":"chat","base_model":"stablelm2:12b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q5_K_S":{"mode":"chat","base_model":"stablelm2:12b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q6_K":{"mode":"chat","base_model":"stablelm2:12b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-q8_0":{"mode":"chat","base_model":"stablelm2:12b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:12b-text":{"mode":"chat","base_model":"stablelm2:12b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:chat":{"mode":"chat","base_model":"stablelm2:chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm2:zephyr":{"mode":"chat","base_model":"stablelm2:zephyr","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b":{"mode":"chat","base_model":"neural-chat:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1":{"mode":"chat","base_model":"neural-chat:7b-v3.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-fp16":{"mode":"chat","base_model":"neural-chat:7b-v3.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q2_K":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q3_K_L":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q3_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q3_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q4_0":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q4_1":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q4_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q4_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q5_0":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q5_1":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q5_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q5_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q6_K":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.1-q8_0":{"mode":"chat","base_model":"neural-chat:7b-v3.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2":{"mode":"chat","base_model":"neural-chat:7b-v3.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-fp16":{"mode":"chat","base_model":"neural-chat:7b-v3.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q2_K":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q3_K_L":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q3_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q3_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q4_0":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q4_1":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q4_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q4_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q5_0":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q5_1":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q5_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q5_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q6_K":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.2-q8_0":{"mode":"chat","base_model":"neural-chat:7b-v3.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3":{"mode":"chat","base_model":"neural-chat:7b-v3.3","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-fp16":{"mode":"chat","base_model":"neural-chat:7b-v3.3-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q2_K":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q3_K_L":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q3_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q3_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q4_0":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q4_1":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q4_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q4_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q5_0":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q5_1":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q5_K_M":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q5_K_S":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q6_K":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"neural-chat:7b-v3.3-q8_0":{"mode":"chat","base_model":"neural-chat:7b-v3.3-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b":{"mode":"chat","base_model":"reflection:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-fp16":{"mode":"chat","base_model":"reflection:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q2_K":{"mode":"chat","base_model":"reflection:70b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q3_K_L":{"mode":"chat","base_model":"reflection:70b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q3_K_M":{"mode":"chat","base_model":"reflection:70b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q3_K_S":{"mode":"chat","base_model":"reflection:70b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q4_0":{"mode":"chat","base_model":"reflection:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q4_1":{"mode":"chat","base_model":"reflection:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q4_K_M":{"mode":"chat","base_model":"reflection:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q4_K_S":{"mode":"chat","base_model":"reflection:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q5_0":{"mode":"chat","base_model":"reflection:70b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q5_1":{"mode":"chat","base_model":"reflection:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q5_K_M":{"mode":"chat","base_model":"reflection:70b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q5_K_S":{"mode":"chat","base_model":"reflection:70b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q6_K":{"mode":"chat","base_model":"reflection:70b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reflection:70b-q8_0":{"mode":"chat","base_model":"reflection:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b":{"mode":"chat","base_model":"wizard-math:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b":{"mode":"chat","base_model":"wizard-math:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b":{"mode":"chat","base_model":"wizard-math:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-fp16":{"mode":"chat","base_model":"wizard-math:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q2_K":{"mode":"chat","base_model":"wizard-math:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q3_K_L":{"mode":"chat","base_model":"wizard-math:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q3_K_M":{"mode":"chat","base_model":"wizard-math:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q3_K_S":{"mode":"chat","base_model":"wizard-math:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q4_0":{"mode":"chat","base_model":"wizard-math:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q4_1":{"mode":"chat","base_model":"wizard-math:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q4_K_M":{"mode":"chat","base_model":"wizard-math:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q4_K_S":{"mode":"chat","base_model":"wizard-math:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q5_0":{"mode":"chat","base_model":"wizard-math:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q5_1":{"mode":"chat","base_model":"wizard-math:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q5_K_M":{"mode":"chat","base_model":"wizard-math:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q5_K_S":{"mode":"chat","base_model":"wizard-math:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q6_K":{"mode":"chat","base_model":"wizard-math:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:13b-q8_0":{"mode":"chat","base_model":"wizard-math:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-fp16":{"mode":"chat","base_model":"wizard-math:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q2_K":{"mode":"chat","base_model":"wizard-math:70b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q3_K_L":{"mode":"chat","base_model":"wizard-math:70b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q3_K_M":{"mode":"chat","base_model":"wizard-math:70b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q3_K_S":{"mode":"chat","base_model":"wizard-math:70b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q4_0":{"mode":"chat","base_model":"wizard-math:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q4_1":{"mode":"chat","base_model":"wizard-math:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q4_K_M":{"mode":"chat","base_model":"wizard-math:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q4_K_S":{"mode":"chat","base_model":"wizard-math:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q5_0":{"mode":"chat","base_model":"wizard-math:70b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q5_1":{"mode":"chat","base_model":"wizard-math:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q5_K_M":{"mode":"chat","base_model":"wizard-math:70b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q5_K_S":{"mode":"chat","base_model":"wizard-math:70b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q6_K":{"mode":"chat","base_model":"wizard-math:70b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:70b-q8_0":{"mode":"chat","base_model":"wizard-math:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-fp16":{"mode":"chat","base_model":"wizard-math:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q2_K":{"mode":"chat","base_model":"wizard-math:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q3_K_L":{"mode":"chat","base_model":"wizard-math:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q3_K_M":{"mode":"chat","base_model":"wizard-math:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q3_K_S":{"mode":"chat","base_model":"wizard-math:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q4_0":{"mode":"chat","base_model":"wizard-math:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q4_1":{"mode":"chat","base_model":"wizard-math:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q4_K_M":{"mode":"chat","base_model":"wizard-math:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q4_K_S":{"mode":"chat","base_model":"wizard-math:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q5_0":{"mode":"chat","base_model":"wizard-math:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q5_1":{"mode":"chat","base_model":"wizard-math:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q5_K_M":{"mode":"chat","base_model":"wizard-math:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q5_K_S":{"mode":"chat","base_model":"wizard-math:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q6_K":{"mode":"chat","base_model":"wizard-math:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-q8_0":{"mode":"chat","base_model":"wizard-math:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-fp16":{"mode":"chat","base_model":"wizard-math:7b-v1.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q2_K":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q3_K_L":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q3_K_M":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q3_K_S":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q4_0":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q4_1":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q4_K_M":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q4_K_S":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q5_0":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q5_1":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q5_K_M":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q5_K_S":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q6_K":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-math:7b-v1.1-q8_0":{"mode":"chat","base_model":"wizard-math:7b-v1.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:1048k":{"mode":"chat","base_model":"llama3-gradient:1048k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b":{"mode":"chat","base_model":"llama3-gradient:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b":{"mode":"chat","base_model":"llama3-gradient:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-fp16":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q2_K":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q3_K_L":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q3_K_M":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q3_K_S":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q4_0":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q4_1":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q4_K_M":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q4_K_S":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q5_0":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q5_1":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q5_K_M":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q5_K_S":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q6_K":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:70b-instruct-1048k-q8_0":{"mode":"chat","base_model":"llama3-gradient:70b-instruct-1048k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-fp16":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q2_K":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q3_K_L":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q3_K_M":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q3_K_S":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q4_0":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q4_1":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q4_K_M":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q4_K_S":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q5_0":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q5_1":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q5_K_M":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q5_K_S":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q6_K":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:8b-instruct-1048k-q8_0":{"mode":"chat","base_model":"llama3-gradient:8b-instruct-1048k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-gradient:instruct":{"mode":"chat","base_model":"llama3-gradient:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b":{"mode":"chat","base_model":"llama3-chatqa:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b":{"mode":"chat","base_model":"llama3-chatqa:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-fp16":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q2_K":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q3_K_L":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q3_K_M":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q3_K_S":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q4_0":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q4_1":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q4_K_M":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q4_K_S":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q5_0":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q5_1":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q5_K_M":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q5_K_S":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q6_K":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:70b-v1.5-q8_0":{"mode":"chat","base_model":"llama3-chatqa:70b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-fp16":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q2_K":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q3_K_L":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q3_K_M":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q3_K_S":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q4_0":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q4_1":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q4_K_M":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q4_K_S":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q5_0":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q5_1":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q5_K_M":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q5_K_S":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q6_K":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-chatqa:8b-v1.5-q8_0":{"mode":"chat","base_model":"llama3-chatqa:8b-v1.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b":{"mode":"chat","base_model":"sqlcoder:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b":{"mode":"chat","base_model":"sqlcoder:15b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-fp16":{"mode":"chat","base_model":"sqlcoder:15b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q2_K":{"mode":"chat","base_model":"sqlcoder:15b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q3_K_L":{"mode":"chat","base_model":"sqlcoder:15b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q3_K_M":{"mode":"chat","base_model":"sqlcoder:15b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q3_K_S":{"mode":"chat","base_model":"sqlcoder:15b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q4_0":{"mode":"chat","base_model":"sqlcoder:15b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q4_1":{"mode":"chat","base_model":"sqlcoder:15b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q4_K_M":{"mode":"chat","base_model":"sqlcoder:15b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q4_K_S":{"mode":"chat","base_model":"sqlcoder:15b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q5_0":{"mode":"chat","base_model":"sqlcoder:15b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q5_1":{"mode":"chat","base_model":"sqlcoder:15b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q5_K_M":{"mode":"chat","base_model":"sqlcoder:15b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q5_K_S":{"mode":"chat","base_model":"sqlcoder:15b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q6_K":{"mode":"chat","base_model":"sqlcoder:15b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:15b-q8_0":{"mode":"chat","base_model":"sqlcoder:15b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-fp16":{"mode":"chat","base_model":"sqlcoder:70b-alpha-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q2_K":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q3_K_L":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q3_K_M":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q3_K_S":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q4_0":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q4_1":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q4_K_M":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q4_K_S":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q5_0":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q5_1":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q5_K_M":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q5_K_S":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q6_K":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:70b-alpha-q8_0":{"mode":"chat","base_model":"sqlcoder:70b-alpha-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-fp16":{"mode":"chat","base_model":"sqlcoder:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q2_K":{"mode":"chat","base_model":"sqlcoder:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q3_K_L":{"mode":"chat","base_model":"sqlcoder:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q3_K_M":{"mode":"chat","base_model":"sqlcoder:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q3_K_S":{"mode":"chat","base_model":"sqlcoder:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q4_0":{"mode":"chat","base_model":"sqlcoder:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q4_1":{"mode":"chat","base_model":"sqlcoder:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q4_K_M":{"mode":"chat","base_model":"sqlcoder:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q4_K_S":{"mode":"chat","base_model":"sqlcoder:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q5_0":{"mode":"chat","base_model":"sqlcoder:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q5_1":{"mode":"chat","base_model":"sqlcoder:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q5_K_M":{"mode":"chat","base_model":"sqlcoder:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q5_K_S":{"mode":"chat","base_model":"sqlcoder:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q6_K":{"mode":"chat","base_model":"sqlcoder:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sqlcoder:7b-q8_0":{"mode":"chat","base_model":"sqlcoder:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bge-large:335m":{"mode":"chat","base_model":"bge-large:335m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bge-large:335m-en-v1.5-fp16":{"mode":"chat","base_model":"bge-large:335m-en-v1.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b":{"mode":"chat","base_model":"xwinlm:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b":{"mode":"chat","base_model":"xwinlm:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1":{"mode":"chat","base_model":"xwinlm:13b-v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-fp16":{"mode":"chat","base_model":"xwinlm:13b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q2_K":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q3_K_L":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q3_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q3_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q4_0":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q4_1":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q4_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q4_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q5_0":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q5_1":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q5_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q5_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q6_K":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.1-q8_0":{"mode":"chat","base_model":"xwinlm:13b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2":{"mode":"chat","base_model":"xwinlm:13b-v0.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-fp16":{"mode":"chat","base_model":"xwinlm:13b-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q2_K":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q3_K_L":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q3_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q3_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q4_0":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q4_1":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q4_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q4_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q5_0":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q5_1":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q5_K_M":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q5_K_S":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q6_K":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:13b-v0.2-q8_0":{"mode":"chat","base_model":"xwinlm:13b-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1":{"mode":"chat","base_model":"xwinlm:70b-v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-fp16":{"mode":"chat","base_model":"xwinlm:70b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q2_K":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q3_K_L":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q3_K_M":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q3_K_S":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q4_0":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q4_1":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q4_K_M":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q4_K_S":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q5_0":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q5_1":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q5_K_S":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q6_K":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:70b-v0.1-q8_0":{"mode":"chat","base_model":"xwinlm:70b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1":{"mode":"chat","base_model":"xwinlm:7b-v0.1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-fp16":{"mode":"chat","base_model":"xwinlm:7b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q2_K":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q3_K_L":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q3_K_M":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q3_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q4_0":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q4_1":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q4_K_M":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q4_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q5_0":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q5_1":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q5_K_M":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q5_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q6_K":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.1-q8_0":{"mode":"chat","base_model":"xwinlm:7b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2":{"mode":"chat","base_model":"xwinlm:7b-v0.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-fp16":{"mode":"chat","base_model":"xwinlm:7b-v0.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q2_K":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q3_K_L":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q3_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q4_0":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q4_1":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q4_K_M":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q4_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q5_0":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q5_K_M":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q5_K_S":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q6_K":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"xwinlm:7b-v0.2-q8_0":{"mode":"chat","base_model":"xwinlm:7b-v0.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b":{"mode":"chat","base_model":"dolphincoder:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b":{"mode":"chat","base_model":"dolphincoder:15b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-fp16":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q2_K":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q3_K_L":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q3_K_M":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q3_K_S":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q4_0":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q4_1":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q4_K_M":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q4_K_S":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q5_0":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q5_1":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q5_K_M":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q5_K_S":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q6_K":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:15b-starcoder2-q8_0":{"mode":"chat","base_model":"dolphincoder:15b-starcoder2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-fp16":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q2_K":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q3_K_L":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q3_K_M":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q3_K_S":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q4_0":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q4_1":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q4_K_M":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q4_K_S":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q5_0":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q5_1":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q5_K_M":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q5_K_S":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q6_K":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphincoder:7b-starcoder2-q8_0":{"mode":"chat","base_model":"dolphincoder:7b-starcoder2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b":{"mode":"chat","base_model":"nous-hermes:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b":{"mode":"chat","base_model":"nous-hermes:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-fp16":{"mode":"chat","base_model":"nous-hermes:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2":{"mode":"chat","base_model":"nous-hermes:13b-llama2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-fp16":{"mode":"chat","base_model":"nous-hermes:13b-llama2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q2_K":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q3_K_L":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q3_K_M":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q3_K_S":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q4_0":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q4_1":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q4_K_M":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q4_K_S":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q5_0":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q5_1":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q5_K_M":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q5_K_S":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q6_K":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-llama2-q8_0":{"mode":"chat","base_model":"nous-hermes:13b-llama2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q2_K":{"mode":"chat","base_model":"nous-hermes:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q3_K_L":{"mode":"chat","base_model":"nous-hermes:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q3_K_M":{"mode":"chat","base_model":"nous-hermes:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q3_K_S":{"mode":"chat","base_model":"nous-hermes:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q4_0":{"mode":"chat","base_model":"nous-hermes:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q4_1":{"mode":"chat","base_model":"nous-hermes:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q4_K_M":{"mode":"chat","base_model":"nous-hermes:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q4_K_S":{"mode":"chat","base_model":"nous-hermes:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q5_0":{"mode":"chat","base_model":"nous-hermes:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q5_1":{"mode":"chat","base_model":"nous-hermes:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q5_K_M":{"mode":"chat","base_model":"nous-hermes:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q5_K_S":{"mode":"chat","base_model":"nous-hermes:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q6_K":{"mode":"chat","base_model":"nous-hermes:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:13b-q8_0":{"mode":"chat","base_model":"nous-hermes:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-fp16":{"mode":"chat","base_model":"nous-hermes:70b-llama2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q2_K":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q3_K_L":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q3_K_M":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q3_K_S":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q4_0":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q4_1":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q4_K_M":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q4_K_S":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q5_0":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q5_1":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q5_K_M":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:70b-llama2-q6_K":{"mode":"chat","base_model":"nous-hermes:70b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2":{"mode":"chat","base_model":"nous-hermes:7b-llama2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-fp16":{"mode":"chat","base_model":"nous-hermes:7b-llama2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q2_K":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q3_K_L":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q3_K_M":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q3_K_S":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q4_0":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q4_1":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q4_K_M":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q4_K_S":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q5_0":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q5_1":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q5_K_M":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q5_K_S":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q6_K":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes:7b-llama2-q8_0":{"mode":"chat","base_model":"nous-hermes:7b-llama2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b":{"mode":"chat","base_model":"phind-codellama:34b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-fp16":{"mode":"chat","base_model":"phind-codellama:34b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python":{"mode":"chat","base_model":"phind-codellama:34b-python","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-fp16":{"mode":"chat","base_model":"phind-codellama:34b-python-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q2_K":{"mode":"chat","base_model":"phind-codellama:34b-python-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q3_K_L":{"mode":"chat","base_model":"phind-codellama:34b-python-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q3_K_M":{"mode":"chat","base_model":"phind-codellama:34b-python-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q3_K_S":{"mode":"chat","base_model":"phind-codellama:34b-python-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q4_0":{"mode":"chat","base_model":"phind-codellama:34b-python-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q4_1":{"mode":"chat","base_model":"phind-codellama:34b-python-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q4_K_M":{"mode":"chat","base_model":"phind-codellama:34b-python-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q4_K_S":{"mode":"chat","base_model":"phind-codellama:34b-python-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q5_0":{"mode":"chat","base_model":"phind-codellama:34b-python-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q5_1":{"mode":"chat","base_model":"phind-codellama:34b-python-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q5_K_M":{"mode":"chat","base_model":"phind-codellama:34b-python-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q5_K_S":{"mode":"chat","base_model":"phind-codellama:34b-python-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q6_K":{"mode":"chat","base_model":"phind-codellama:34b-python-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-python-q8_0":{"mode":"chat","base_model":"phind-codellama:34b-python-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q2_K":{"mode":"chat","base_model":"phind-codellama:34b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q3_K_L":{"mode":"chat","base_model":"phind-codellama:34b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q3_K_M":{"mode":"chat","base_model":"phind-codellama:34b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q3_K_S":{"mode":"chat","base_model":"phind-codellama:34b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q4_0":{"mode":"chat","base_model":"phind-codellama:34b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q4_1":{"mode":"chat","base_model":"phind-codellama:34b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q4_K_M":{"mode":"chat","base_model":"phind-codellama:34b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q4_K_S":{"mode":"chat","base_model":"phind-codellama:34b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q5_0":{"mode":"chat","base_model":"phind-codellama:34b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q5_1":{"mode":"chat","base_model":"phind-codellama:34b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q5_K_M":{"mode":"chat","base_model":"phind-codellama:34b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q5_K_S":{"mode":"chat","base_model":"phind-codellama:34b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q6_K":{"mode":"chat","base_model":"phind-codellama:34b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-q8_0":{"mode":"chat","base_model":"phind-codellama:34b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-fp16":{"mode":"chat","base_model":"phind-codellama:34b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q2_K":{"mode":"chat","base_model":"phind-codellama:34b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q3_K_L":{"mode":"chat","base_model":"phind-codellama:34b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q3_K_M":{"mode":"chat","base_model":"phind-codellama:34b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q3_K_S":{"mode":"chat","base_model":"phind-codellama:34b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q4_0":{"mode":"chat","base_model":"phind-codellama:34b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q4_1":{"mode":"chat","base_model":"phind-codellama:34b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q4_K_M":{"mode":"chat","base_model":"phind-codellama:34b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q4_K_S":{"mode":"chat","base_model":"phind-codellama:34b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q5_0":{"mode":"chat","base_model":"phind-codellama:34b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q5_1":{"mode":"chat","base_model":"phind-codellama:34b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q5_K_M":{"mode":"chat","base_model":"phind-codellama:34b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q5_K_S":{"mode":"chat","base_model":"phind-codellama:34b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q6_K":{"mode":"chat","base_model":"phind-codellama:34b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phind-codellama:34b-v2-q8_0":{"mode":"chat","base_model":"phind-codellama:34b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-phi3:3.8b":{"mode":"chat","base_model":"llava-phi3:3.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-phi3:3.8b-mini-fp16":{"mode":"chat","base_model":"llava-phi3:3.8b-mini-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llava-phi3:3.8b-mini-q4_0":{"mode":"chat","base_model":"llava-phi3:3.8b-mini-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b":{"mode":"chat","base_model":"yarn-llama2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b":{"mode":"chat","base_model":"yarn-llama2:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k":{"mode":"chat","base_model":"yarn-llama2:13b-128k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-fp16":{"mode":"chat","base_model":"yarn-llama2:13b-128k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q2_K":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q3_K_L":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q3_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q3_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q4_0":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q4_1":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q4_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q4_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q5_0":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q5_1":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q5_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q5_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q6_K":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-128k-q8_0":{"mode":"chat","base_model":"yarn-llama2:13b-128k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k":{"mode":"chat","base_model":"yarn-llama2:13b-64k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-fp16":{"mode":"chat","base_model":"yarn-llama2:13b-64k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q2_K":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q3_K_L":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q3_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q3_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q4_0":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q4_1":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q4_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q4_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q5_0":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q5_1":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q5_K_M":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q5_K_S":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q6_K":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:13b-64k-q8_0":{"mode":"chat","base_model":"yarn-llama2:13b-64k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k":{"mode":"chat","base_model":"yarn-llama2:7b-128k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-fp16":{"mode":"chat","base_model":"yarn-llama2:7b-128k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q2_K":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q3_K_L":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q3_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q3_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q4_0":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q4_1":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q4_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q4_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q5_0":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q5_1":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q5_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q5_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q6_K":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-128k-q8_0":{"mode":"chat","base_model":"yarn-llama2:7b-128k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k":{"mode":"chat","base_model":"yarn-llama2:7b-64k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-fp16":{"mode":"chat","base_model":"yarn-llama2:7b-64k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q2_K":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q3_K_L":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q3_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q3_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q4_0":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q4_1":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q4_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q4_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q5_0":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q5_1":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q5_K_M":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q5_K_S":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q6_K":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-llama2:7b-64k-q8_0":{"mode":"chat","base_model":"yarn-llama2:7b-64k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b":{"mode":"chat","base_model":"solar:10.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-fp16":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q2_K":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q3_K_L":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q3_K_M":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q3_K_S":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q4_0":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q4_1":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q4_K_M":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q4_K_S":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q5_0":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q5_1":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q5_K_M":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q5_K_S":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q6_K":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-instruct-v1-q8_0":{"mode":"chat","base_model":"solar:10.7b-instruct-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-fp16":{"mode":"chat","base_model":"solar:10.7b-text-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q2_K":{"mode":"chat","base_model":"solar:10.7b-text-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q3_K_L":{"mode":"chat","base_model":"solar:10.7b-text-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q3_K_M":{"mode":"chat","base_model":"solar:10.7b-text-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q3_K_S":{"mode":"chat","base_model":"solar:10.7b-text-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q4_0":{"mode":"chat","base_model":"solar:10.7b-text-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q4_1":{"mode":"chat","base_model":"solar:10.7b-text-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q4_K_M":{"mode":"chat","base_model":"solar:10.7b-text-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q4_K_S":{"mode":"chat","base_model":"solar:10.7b-text-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q5_0":{"mode":"chat","base_model":"solar:10.7b-text-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q5_1":{"mode":"chat","base_model":"solar:10.7b-text-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q5_K_M":{"mode":"chat","base_model":"solar:10.7b-text-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q5_K_S":{"mode":"chat","base_model":"solar:10.7b-text-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q6_K":{"mode":"chat","base_model":"solar:10.7b-text-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar:10.7b-text-v1-q8_0":{"mode":"chat","base_model":"solar:10.7b-text-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b":{"mode":"chat","base_model":"granite3.1-dense:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b":{"mode":"chat","base_model":"granite3.1-dense:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-fp16":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q2_K":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q4_0":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q4_1":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q5_0":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q5_1":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q6_K":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:2b-instruct-q8_0":{"mode":"chat","base_model":"granite3.1-dense:2b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-fp16":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q2_K":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q4_0":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q4_1":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q5_0":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q5_1":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q6_K":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-dense:8b-instruct-q8_0":{"mode":"chat","base_model":"granite3.1-dense:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b":{"mode":"chat","base_model":"starling-lm:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha":{"mode":"chat","base_model":"starling-lm:7b-alpha","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-fp16":{"mode":"chat","base_model":"starling-lm:7b-alpha-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q2_K":{"mode":"chat","base_model":"starling-lm:7b-alpha-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q3_K_L":{"mode":"chat","base_model":"starling-lm:7b-alpha-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q3_K_M":{"mode":"chat","base_model":"starling-lm:7b-alpha-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q3_K_S":{"mode":"chat","base_model":"starling-lm:7b-alpha-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q4_0":{"mode":"chat","base_model":"starling-lm:7b-alpha-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q4_1":{"mode":"chat","base_model":"starling-lm:7b-alpha-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q4_K_M":{"mode":"chat","base_model":"starling-lm:7b-alpha-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q4_K_S":{"mode":"chat","base_model":"starling-lm:7b-alpha-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q5_0":{"mode":"chat","base_model":"starling-lm:7b-alpha-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q5_1":{"mode":"chat","base_model":"starling-lm:7b-alpha-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q5_K_M":{"mode":"chat","base_model":"starling-lm:7b-alpha-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q5_K_S":{"mode":"chat","base_model":"starling-lm:7b-alpha-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q6_K":{"mode":"chat","base_model":"starling-lm:7b-alpha-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-alpha-q8_0":{"mode":"chat","base_model":"starling-lm:7b-alpha-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta":{"mode":"chat","base_model":"starling-lm:7b-beta","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-fp16":{"mode":"chat","base_model":"starling-lm:7b-beta-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q2_K":{"mode":"chat","base_model":"starling-lm:7b-beta-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q3_K_L":{"mode":"chat","base_model":"starling-lm:7b-beta-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q3_K_M":{"mode":"chat","base_model":"starling-lm:7b-beta-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q3_K_S":{"mode":"chat","base_model":"starling-lm:7b-beta-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q4_0":{"mode":"chat","base_model":"starling-lm:7b-beta-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q4_1":{"mode":"chat","base_model":"starling-lm:7b-beta-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q4_K_M":{"mode":"chat","base_model":"starling-lm:7b-beta-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q4_K_S":{"mode":"chat","base_model":"starling-lm:7b-beta-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q5_0":{"mode":"chat","base_model":"starling-lm:7b-beta-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q5_1":{"mode":"chat","base_model":"starling-lm:7b-beta-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q5_K_M":{"mode":"chat","base_model":"starling-lm:7b-beta-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q5_K_S":{"mode":"chat","base_model":"starling-lm:7b-beta-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q6_K":{"mode":"chat","base_model":"starling-lm:7b-beta-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:7b-beta-q8_0":{"mode":"chat","base_model":"starling-lm:7b-beta-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:alpha":{"mode":"chat","base_model":"starling-lm:alpha","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"starling-lm:beta":{"mode":"chat","base_model":"starling-lm:beta","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b":{"mode":"chat","base_model":"athene-v2:72b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-fp16":{"mode":"chat","base_model":"athene-v2:72b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q2_K":{"mode":"chat","base_model":"athene-v2:72b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q3_K_L":{"mode":"chat","base_model":"athene-v2:72b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q3_K_M":{"mode":"chat","base_model":"athene-v2:72b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q3_K_S":{"mode":"chat","base_model":"athene-v2:72b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q4_0":{"mode":"chat","base_model":"athene-v2:72b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q4_1":{"mode":"chat","base_model":"athene-v2:72b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q4_K_M":{"mode":"chat","base_model":"athene-v2:72b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q4_K_S":{"mode":"chat","base_model":"athene-v2:72b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q5_0":{"mode":"chat","base_model":"athene-v2:72b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q5_1":{"mode":"chat","base_model":"athene-v2:72b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q5_K_M":{"mode":"chat","base_model":"athene-v2:72b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q5_K_S":{"mode":"chat","base_model":"athene-v2:72b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q6_K":{"mode":"chat","base_model":"athene-v2:72b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"athene-v2:72b-q8_0":{"mode":"chat","base_model":"athene-v2:72b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b":{"mode":"chat","base_model":"yi-coder:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b":{"mode":"chat","base_model":"yi-coder:9b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base":{"mode":"chat","base_model":"yi-coder:1.5b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-fp16":{"mode":"chat","base_model":"yi-coder:1.5b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q2_K":{"mode":"chat","base_model":"yi-coder:1.5b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q3_K_L":{"mode":"chat","base_model":"yi-coder:1.5b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q3_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q3_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q4_0":{"mode":"chat","base_model":"yi-coder:1.5b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q4_1":{"mode":"chat","base_model":"yi-coder:1.5b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q4_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q4_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q5_0":{"mode":"chat","base_model":"yi-coder:1.5b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q5_1":{"mode":"chat","base_model":"yi-coder:1.5b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q5_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q5_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q6_K":{"mode":"chat","base_model":"yi-coder:1.5b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-base-q8_0":{"mode":"chat","base_model":"yi-coder:1.5b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat":{"mode":"chat","base_model":"yi-coder:1.5b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-fp16":{"mode":"chat","base_model":"yi-coder:1.5b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q2_K":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q3_K_L":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q3_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q3_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q4_0":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q4_1":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q4_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q4_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q5_0":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q5_1":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q5_K_M":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q5_K_S":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q6_K":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:1.5b-chat-q8_0":{"mode":"chat","base_model":"yi-coder:1.5b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base":{"mode":"chat","base_model":"yi-coder:9b-base","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-fp16":{"mode":"chat","base_model":"yi-coder:9b-base-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q2_K":{"mode":"chat","base_model":"yi-coder:9b-base-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q3_K_L":{"mode":"chat","base_model":"yi-coder:9b-base-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q3_K_M":{"mode":"chat","base_model":"yi-coder:9b-base-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q3_K_S":{"mode":"chat","base_model":"yi-coder:9b-base-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q4_0":{"mode":"chat","base_model":"yi-coder:9b-base-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q4_1":{"mode":"chat","base_model":"yi-coder:9b-base-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q4_K_M":{"mode":"chat","base_model":"yi-coder:9b-base-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q4_K_S":{"mode":"chat","base_model":"yi-coder:9b-base-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q5_0":{"mode":"chat","base_model":"yi-coder:9b-base-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q5_1":{"mode":"chat","base_model":"yi-coder:9b-base-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q5_K_M":{"mode":"chat","base_model":"yi-coder:9b-base-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q5_K_S":{"mode":"chat","base_model":"yi-coder:9b-base-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q6_K":{"mode":"chat","base_model":"yi-coder:9b-base-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-base-q8_0":{"mode":"chat","base_model":"yi-coder:9b-base-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat":{"mode":"chat","base_model":"yi-coder:9b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-fp16":{"mode":"chat","base_model":"yi-coder:9b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q2_K":{"mode":"chat","base_model":"yi-coder:9b-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q3_K_L":{"mode":"chat","base_model":"yi-coder:9b-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q3_K_M":{"mode":"chat","base_model":"yi-coder:9b-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q3_K_S":{"mode":"chat","base_model":"yi-coder:9b-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q4_0":{"mode":"chat","base_model":"yi-coder:9b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q4_1":{"mode":"chat","base_model":"yi-coder:9b-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q4_K_M":{"mode":"chat","base_model":"yi-coder:9b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q4_K_S":{"mode":"chat","base_model":"yi-coder:9b-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q5_0":{"mode":"chat","base_model":"yi-coder:9b-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q5_1":{"mode":"chat","base_model":"yi-coder:9b-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q5_K_M":{"mode":"chat","base_model":"yi-coder:9b-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q5_K_S":{"mode":"chat","base_model":"yi-coder:9b-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q6_K":{"mode":"chat","base_model":"yi-coder:9b-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yi-coder:9b-chat-q8_0":{"mode":"chat","base_model":"yi-coder:9b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-fp16":{"mode":"chat","base_model":"wizardlm:13b-llama2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q2_K":{"mode":"chat","base_model":"wizardlm:13b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q3_K_L":{"mode":"chat","base_model":"wizardlm:13b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q3_K_M":{"mode":"chat","base_model":"wizardlm:13b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q3_K_S":{"mode":"chat","base_model":"wizardlm:13b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q4_0":{"mode":"chat","base_model":"wizardlm:13b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q4_1":{"mode":"chat","base_model":"wizardlm:13b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q4_K_M":{"mode":"chat","base_model":"wizardlm:13b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q4_K_S":{"mode":"chat","base_model":"wizardlm:13b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q5_0":{"mode":"chat","base_model":"wizardlm:13b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q5_1":{"mode":"chat","base_model":"wizardlm:13b-llama2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q5_K_M":{"mode":"chat","base_model":"wizardlm:13b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q5_K_S":{"mode":"chat","base_model":"wizardlm:13b-llama2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q6_K":{"mode":"chat","base_model":"wizardlm:13b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-llama2-q8_0":{"mode":"chat","base_model":"wizardlm:13b-llama2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q2_K":{"mode":"chat","base_model":"wizardlm:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q3_K_L":{"mode":"chat","base_model":"wizardlm:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q3_K_M":{"mode":"chat","base_model":"wizardlm:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q3_K_S":{"mode":"chat","base_model":"wizardlm:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q4_0":{"mode":"chat","base_model":"wizardlm:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q4_1":{"mode":"chat","base_model":"wizardlm:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q4_K_M":{"mode":"chat","base_model":"wizardlm:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q4_K_S":{"mode":"chat","base_model":"wizardlm:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q5_0":{"mode":"chat","base_model":"wizardlm:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q5_1":{"mode":"chat","base_model":"wizardlm:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q5_K_M":{"mode":"chat","base_model":"wizardlm:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q5_K_S":{"mode":"chat","base_model":"wizardlm:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q6_K":{"mode":"chat","base_model":"wizardlm:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:13b-q8_0":{"mode":"chat","base_model":"wizardlm:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-fp16":{"mode":"chat","base_model":"wizardlm:30b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q2_K":{"mode":"chat","base_model":"wizardlm:30b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q3_K_L":{"mode":"chat","base_model":"wizardlm:30b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q3_K_M":{"mode":"chat","base_model":"wizardlm:30b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q3_K_S":{"mode":"chat","base_model":"wizardlm:30b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q4_0":{"mode":"chat","base_model":"wizardlm:30b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q4_1":{"mode":"chat","base_model":"wizardlm:30b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q4_K_M":{"mode":"chat","base_model":"wizardlm:30b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q4_K_S":{"mode":"chat","base_model":"wizardlm:30b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q5_0":{"mode":"chat","base_model":"wizardlm:30b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q5_1":{"mode":"chat","base_model":"wizardlm:30b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q5_K_M":{"mode":"chat","base_model":"wizardlm:30b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q5_K_S":{"mode":"chat","base_model":"wizardlm:30b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q6_K":{"mode":"chat","base_model":"wizardlm:30b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:30b-q8_0":{"mode":"chat","base_model":"wizardlm:30b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q2_K":{"mode":"chat","base_model":"wizardlm:70b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q3_K_L":{"mode":"chat","base_model":"wizardlm:70b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q3_K_M":{"mode":"chat","base_model":"wizardlm:70b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q3_K_S":{"mode":"chat","base_model":"wizardlm:70b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q4_0":{"mode":"chat","base_model":"wizardlm:70b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q4_1":{"mode":"chat","base_model":"wizardlm:70b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q4_K_M":{"mode":"chat","base_model":"wizardlm:70b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q4_K_S":{"mode":"chat","base_model":"wizardlm:70b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q5_0":{"mode":"chat","base_model":"wizardlm:70b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q5_K_M":{"mode":"chat","base_model":"wizardlm:70b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q5_K_S":{"mode":"chat","base_model":"wizardlm:70b-llama2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q6_K":{"mode":"chat","base_model":"wizardlm:70b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:70b-llama2-q8_0":{"mode":"chat","base_model":"wizardlm:70b-llama2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-fp16":{"mode":"chat","base_model":"wizardlm:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q2_K":{"mode":"chat","base_model":"wizardlm:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q3_K_L":{"mode":"chat","base_model":"wizardlm:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q3_K_M":{"mode":"chat","base_model":"wizardlm:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q3_K_S":{"mode":"chat","base_model":"wizardlm:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q4_0":{"mode":"chat","base_model":"wizardlm:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q4_1":{"mode":"chat","base_model":"wizardlm:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q4_K_M":{"mode":"chat","base_model":"wizardlm:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q4_K_S":{"mode":"chat","base_model":"wizardlm:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q5_0":{"mode":"chat","base_model":"wizardlm:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q5_1":{"mode":"chat","base_model":"wizardlm:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q5_K_M":{"mode":"chat","base_model":"wizardlm:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q5_K_S":{"mode":"chat","base_model":"wizardlm:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q6_K":{"mode":"chat","base_model":"wizardlm:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm:7b-q8_0":{"mode":"chat","base_model":"wizardlm:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b":{"mode":"chat","base_model":"samantha-mistral:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-fp16":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q2_K":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q3_K_L":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q3_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q3_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q4_0":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q4_1":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q4_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q4_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q5_0":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q5_1":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q5_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q5_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q6_K":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-instruct-q8_0":{"mode":"chat","base_model":"samantha-mistral:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text":{"mode":"chat","base_model":"samantha-mistral:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-fp16":{"mode":"chat","base_model":"samantha-mistral:7b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q2_K":{"mode":"chat","base_model":"samantha-mistral:7b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q3_K_L":{"mode":"chat","base_model":"samantha-mistral:7b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q3_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q3_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q4_0":{"mode":"chat","base_model":"samantha-mistral:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q4_1":{"mode":"chat","base_model":"samantha-mistral:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q4_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q4_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q5_0":{"mode":"chat","base_model":"samantha-mistral:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q5_1":{"mode":"chat","base_model":"samantha-mistral:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q5_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q5_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q6_K":{"mode":"chat","base_model":"samantha-mistral:7b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-text-q8_0":{"mode":"chat","base_model":"samantha-mistral:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-fp16":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q2_K":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q3_K_L":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q3_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q3_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q4_0":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q4_1":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q4_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q4_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q5_0":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q5_1":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q5_K_M":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q5_K_S":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q6_K":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"samantha-mistral:7b-v1.2-text-q8_0":{"mode":"chat","base_model":"samantha-mistral:7b-v1.2-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1m":{"mode":"chat","base_model":"internlm2:1m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b":{"mode":"chat","base_model":"internlm2:1.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b":{"mode":"chat","base_model":"internlm2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b":{"mode":"chat","base_model":"internlm2:20b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-fp16":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q2_K":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q3_K_L":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q3_K_M":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q3_K_S":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q4_0":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q4_1":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q4_K_M":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q4_K_S":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q5_0":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q5_1":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q5_K_M":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q5_K_S":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q6_K":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:1.8b-chat-v2.5-q8_0":{"mode":"chat","base_model":"internlm2:1.8b-chat-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-fp16":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q2_K":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q3_K_L":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q3_K_M":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q3_K_S":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q4_0":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q4_1":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q4_K_M":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q4_K_S":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q5_0":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q5_1":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q5_K_M":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q5_K_S":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q6_K":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:20b-chat-v2.5-q8_0":{"mode":"chat","base_model":"internlm2:20b-chat-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-fp16":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q2_K":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q3_K_L":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q3_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q3_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q4_0":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q4_1":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q4_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q4_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q5_0":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q5_1":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q5_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q5_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q6_K":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-1m-v2.5-q8_0":{"mode":"chat","base_model":"internlm2:7b-chat-1m-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-fp16":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q2_K":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q3_K_L":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q3_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q3_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q4_0":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q4_1":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q4_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q4_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q5_0":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q5_1":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q5_K_M":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q5_K_S":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q6_K":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"internlm2:7b-chat-v2.5-q8_0":{"mode":"chat","base_model":"internlm2:7b-chat-v2.5-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b":{"mode":"chat","base_model":"falcon:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b":{"mode":"chat","base_model":"falcon:40b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:180b":{"mode":"chat","base_model":"falcon:180b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:180b-chat":{"mode":"chat","base_model":"falcon:180b-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:180b-chat-q4_0":{"mode":"chat","base_model":"falcon:180b-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:180b-text":{"mode":"chat","base_model":"falcon:180b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:180b-text-q4_0":{"mode":"chat","base_model":"falcon:180b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct":{"mode":"chat","base_model":"falcon:40b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-fp16":{"mode":"chat","base_model":"falcon:40b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-q4_0":{"mode":"chat","base_model":"falcon:40b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-q4_1":{"mode":"chat","base_model":"falcon:40b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-q5_0":{"mode":"chat","base_model":"falcon:40b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-q5_1":{"mode":"chat","base_model":"falcon:40b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-instruct-q8_0":{"mode":"chat","base_model":"falcon:40b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text":{"mode":"chat","base_model":"falcon:40b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-fp16":{"mode":"chat","base_model":"falcon:40b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-q4_0":{"mode":"chat","base_model":"falcon:40b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-q4_1":{"mode":"chat","base_model":"falcon:40b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-q5_0":{"mode":"chat","base_model":"falcon:40b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-q5_1":{"mode":"chat","base_model":"falcon:40b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:40b-text-q8_0":{"mode":"chat","base_model":"falcon:40b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct":{"mode":"chat","base_model":"falcon:7b-instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-fp16":{"mode":"chat","base_model":"falcon:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-q4_0":{"mode":"chat","base_model":"falcon:7b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-q4_1":{"mode":"chat","base_model":"falcon:7b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-q5_0":{"mode":"chat","base_model":"falcon:7b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-q5_1":{"mode":"chat","base_model":"falcon:7b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-instruct-q8_0":{"mode":"chat","base_model":"falcon:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text":{"mode":"chat","base_model":"falcon:7b-text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-fp16":{"mode":"chat","base_model":"falcon:7b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-q4_0":{"mode":"chat","base_model":"falcon:7b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-q4_1":{"mode":"chat","base_model":"falcon:7b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-q5_0":{"mode":"chat","base_model":"falcon:7b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-q5_1":{"mode":"chat","base_model":"falcon:7b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:7b-text-q8_0":{"mode":"chat","base_model":"falcon:7b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:instruct":{"mode":"chat","base_model":"falcon:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon:text":{"mode":"chat","base_model":"falcon:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b":{"mode":"chat","base_model":"nemotron-mini:4b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-fp16":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q2_K":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q3_K_L":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q3_K_M":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q3_K_S":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q4_0":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q4_1":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q4_K_M":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q4_K_S":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q5_0":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q5_1":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q5_K_M":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q5_K_S":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q6_K":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron-mini:4b-instruct-q8_0":{"mode":"chat","base_model":"nemotron-mini:4b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b":{"mode":"chat","base_model":"nemotron:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-fp16":{"mode":"chat","base_model":"nemotron:70b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q2_K":{"mode":"chat","base_model":"nemotron:70b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q3_K_L":{"mode":"chat","base_model":"nemotron:70b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q3_K_M":{"mode":"chat","base_model":"nemotron:70b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q3_K_S":{"mode":"chat","base_model":"nemotron:70b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q4_0":{"mode":"chat","base_model":"nemotron:70b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q4_1":{"mode":"chat","base_model":"nemotron:70b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q4_K_M":{"mode":"chat","base_model":"nemotron:70b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q4_K_S":{"mode":"chat","base_model":"nemotron:70b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q5_0":{"mode":"chat","base_model":"nemotron:70b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q5_1":{"mode":"chat","base_model":"nemotron:70b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q5_K_M":{"mode":"chat","base_model":"nemotron:70b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q5_K_S":{"mode":"chat","base_model":"nemotron:70b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q6_K":{"mode":"chat","base_model":"nemotron:70b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nemotron:70b-instruct-q8_0":{"mode":"chat","base_model":"nemotron:70b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b":{"mode":"chat","base_model":"dolphin-phi:2.7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q2_K":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q3_K_L":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q3_K_M":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q3_K_S":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q4_0":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q4_K_M":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q4_K_S":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q5_0":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q5_K_M":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q5_K_S":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q6_K":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dolphin-phi:2.7b-v2.6-q8_0":{"mode":"chat","base_model":"dolphin-phi:2.7b-v2.6-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b":{"mode":"chat","base_model":"orca2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b":{"mode":"chat","base_model":"orca2:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-fp16":{"mode":"chat","base_model":"orca2:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q2_K":{"mode":"chat","base_model":"orca2:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q3_K_L":{"mode":"chat","base_model":"orca2:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q3_K_M":{"mode":"chat","base_model":"orca2:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q3_K_S":{"mode":"chat","base_model":"orca2:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q4_0":{"mode":"chat","base_model":"orca2:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q4_1":{"mode":"chat","base_model":"orca2:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q4_K_M":{"mode":"chat","base_model":"orca2:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q4_K_S":{"mode":"chat","base_model":"orca2:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q5_0":{"mode":"chat","base_model":"orca2:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q5_1":{"mode":"chat","base_model":"orca2:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q5_K_M":{"mode":"chat","base_model":"orca2:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q5_K_S":{"mode":"chat","base_model":"orca2:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q6_K":{"mode":"chat","base_model":"orca2:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:13b-q8_0":{"mode":"chat","base_model":"orca2:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-fp16":{"mode":"chat","base_model":"orca2:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q2_K":{"mode":"chat","base_model":"orca2:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q3_K_L":{"mode":"chat","base_model":"orca2:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q3_K_M":{"mode":"chat","base_model":"orca2:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q3_K_S":{"mode":"chat","base_model":"orca2:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q4_0":{"mode":"chat","base_model":"orca2:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q4_1":{"mode":"chat","base_model":"orca2:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q4_K_M":{"mode":"chat","base_model":"orca2:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q4_K_S":{"mode":"chat","base_model":"orca2:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q5_0":{"mode":"chat","base_model":"orca2:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q5_1":{"mode":"chat","base_model":"orca2:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q5_K_M":{"mode":"chat","base_model":"orca2:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q5_K_S":{"mode":"chat","base_model":"orca2:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q6_K":{"mode":"chat","base_model":"orca2:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"orca2:7b-q8_0":{"mode":"chat","base_model":"orca2:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepscaler:1.5b":{"mode":"chat","base_model":"deepscaler:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepscaler:1.5b-preview-fp16":{"mode":"chat","base_model":"deepscaler:1.5b-preview-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepscaler:1.5b-preview-q4_K_M":{"mode":"chat","base_model":"deepscaler:1.5b-preview-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepscaler:1.5b-preview-q8_0":{"mode":"chat","base_model":"deepscaler:1.5b-preview-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b":{"mode":"chat","base_model":"wizardlm-uncensored:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-fp16":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q2_K":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q3_K_L":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q3_K_M":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q3_K_S":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q4_0":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q4_1":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q4_K_M":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q4_K_S":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q5_0":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q5_1":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q5_K_M":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q5_K_S":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q6_K":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizardlm-uncensored:13b-llama2-q8_0":{"mode":"chat","base_model":"wizardlm-uncensored:13b-llama2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b":{"mode":"chat","base_model":"stable-beluga:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b":{"mode":"chat","base_model":"stable-beluga:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b":{"mode":"chat","base_model":"stable-beluga:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-fp16":{"mode":"chat","base_model":"stable-beluga:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q2_K":{"mode":"chat","base_model":"stable-beluga:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q3_K_L":{"mode":"chat","base_model":"stable-beluga:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q3_K_M":{"mode":"chat","base_model":"stable-beluga:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q3_K_S":{"mode":"chat","base_model":"stable-beluga:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q4_0":{"mode":"chat","base_model":"stable-beluga:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q4_1":{"mode":"chat","base_model":"stable-beluga:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q4_K_M":{"mode":"chat","base_model":"stable-beluga:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q4_K_S":{"mode":"chat","base_model":"stable-beluga:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q5_0":{"mode":"chat","base_model":"stable-beluga:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q5_1":{"mode":"chat","base_model":"stable-beluga:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q5_K_M":{"mode":"chat","base_model":"stable-beluga:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q5_K_S":{"mode":"chat","base_model":"stable-beluga:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q6_K":{"mode":"chat","base_model":"stable-beluga:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:13b-q8_0":{"mode":"chat","base_model":"stable-beluga:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-fp16":{"mode":"chat","base_model":"stable-beluga:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q2_K":{"mode":"chat","base_model":"stable-beluga:70b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q3_K_L":{"mode":"chat","base_model":"stable-beluga:70b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q3_K_M":{"mode":"chat","base_model":"stable-beluga:70b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q3_K_S":{"mode":"chat","base_model":"stable-beluga:70b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q4_0":{"mode":"chat","base_model":"stable-beluga:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q4_1":{"mode":"chat","base_model":"stable-beluga:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q4_K_M":{"mode":"chat","base_model":"stable-beluga:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q4_K_S":{"mode":"chat","base_model":"stable-beluga:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q5_0":{"mode":"chat","base_model":"stable-beluga:70b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q5_1":{"mode":"chat","base_model":"stable-beluga:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q5_K_M":{"mode":"chat","base_model":"stable-beluga:70b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q5_K_S":{"mode":"chat","base_model":"stable-beluga:70b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q6_K":{"mode":"chat","base_model":"stable-beluga:70b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:70b-q8_0":{"mode":"chat","base_model":"stable-beluga:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-fp16":{"mode":"chat","base_model":"stable-beluga:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q2_K":{"mode":"chat","base_model":"stable-beluga:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q3_K_L":{"mode":"chat","base_model":"stable-beluga:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q3_K_M":{"mode":"chat","base_model":"stable-beluga:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q3_K_S":{"mode":"chat","base_model":"stable-beluga:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q4_0":{"mode":"chat","base_model":"stable-beluga:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q4_1":{"mode":"chat","base_model":"stable-beluga:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q4_K_M":{"mode":"chat","base_model":"stable-beluga:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q4_K_S":{"mode":"chat","base_model":"stable-beluga:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q5_0":{"mode":"chat","base_model":"stable-beluga:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q5_1":{"mode":"chat","base_model":"stable-beluga:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q5_K_M":{"mode":"chat","base_model":"stable-beluga:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q5_K_S":{"mode":"chat","base_model":"stable-beluga:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q6_K":{"mode":"chat","base_model":"stable-beluga:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stable-beluga:7b-q8_0":{"mode":"chat","base_model":"stable-beluga:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b":{"mode":"chat","base_model":"granite3-dense:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b":{"mode":"chat","base_model":"granite3-dense:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-fp16":{"mode":"chat","base_model":"granite3-dense:2b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q2_K":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q4_0":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q4_1":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q5_0":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q5_1":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q6_K":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:2b-instruct-q8_0":{"mode":"chat","base_model":"granite3-dense:2b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-fp16":{"mode":"chat","base_model":"granite3-dense:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q2_K":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q4_0":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q4_1":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q5_0":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q5_1":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q6_K":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-dense:8b-instruct-q8_0":{"mode":"chat","base_model":"granite3-dense:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b":{"mode":"chat","base_model":"llama3-groq-tool-use:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b":{"mode":"chat","base_model":"llama3-groq-tool-use:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-fp16":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q2_K":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q3_K_L":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q3_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q3_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q4_0":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q4_1":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q4_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q4_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q5_0":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q5_1":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q5_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q5_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q6_K":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:70b-q8_0":{"mode":"chat","base_model":"llama3-groq-tool-use:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-fp16":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q2_K":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q3_K_L":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q3_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q3_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q4_0":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q4_1":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q4_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q4_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q5_0":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q5_1":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q5_K_M":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q5_K_S":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q6_K":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama3-groq-tool-use:8b-q8_0":{"mode":"chat","base_model":"llama3-groq-tool-use:8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b":{"mode":"chat","base_model":"deepseek-v2.5:236b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b-q4_0":{"mode":"chat","base_model":"deepseek-v2.5:236b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b-q4_1":{"mode":"chat","base_model":"deepseek-v2.5:236b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b-q5_0":{"mode":"chat","base_model":"deepseek-v2.5:236b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b-q5_1":{"mode":"chat","base_model":"deepseek-v2.5:236b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"deepseek-v2.5:236b-q8_0":{"mode":"chat","base_model":"deepseek-v2.5:236b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b":{"mode":"chat","base_model":"medllama2:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-fp16":{"mode":"chat","base_model":"medllama2:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q2_K":{"mode":"chat","base_model":"medllama2:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q3_K_L":{"mode":"chat","base_model":"medllama2:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q3_K_M":{"mode":"chat","base_model":"medllama2:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q3_K_S":{"mode":"chat","base_model":"medllama2:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q4_0":{"mode":"chat","base_model":"medllama2:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q4_1":{"mode":"chat","base_model":"medllama2:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q4_K_M":{"mode":"chat","base_model":"medllama2:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q4_K_S":{"mode":"chat","base_model":"medllama2:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q5_0":{"mode":"chat","base_model":"medllama2:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q5_1":{"mode":"chat","base_model":"medllama2:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q5_K_M":{"mode":"chat","base_model":"medllama2:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q5_K_S":{"mode":"chat","base_model":"medllama2:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q6_K":{"mode":"chat","base_model":"medllama2:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"medllama2:7b-q8_0":{"mode":"chat","base_model":"medllama2:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b":{"mode":"chat","base_model":"meditron:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:70b":{"mode":"chat","base_model":"meditron:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:70b-q4_0":{"mode":"chat","base_model":"meditron:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:70b-q4_1":{"mode":"chat","base_model":"meditron:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:70b-q4_K_S":{"mode":"chat","base_model":"meditron:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:70b-q5_1":{"mode":"chat","base_model":"meditron:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-fp16":{"mode":"chat","base_model":"meditron:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q2_K":{"mode":"chat","base_model":"meditron:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q3_K_L":{"mode":"chat","base_model":"meditron:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q3_K_M":{"mode":"chat","base_model":"meditron:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q3_K_S":{"mode":"chat","base_model":"meditron:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q4_0":{"mode":"chat","base_model":"meditron:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q4_1":{"mode":"chat","base_model":"meditron:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q4_K_M":{"mode":"chat","base_model":"meditron:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q4_K_S":{"mode":"chat","base_model":"meditron:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q5_0":{"mode":"chat","base_model":"meditron:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q5_1":{"mode":"chat","base_model":"meditron:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q5_K_M":{"mode":"chat","base_model":"meditron:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q5_K_S":{"mode":"chat","base_model":"meditron:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q6_K":{"mode":"chat","base_model":"meditron:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"meditron:7b-q8_0":{"mode":"chat","base_model":"meditron:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smallthinker:3b":{"mode":"chat","base_model":"smallthinker:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smallthinker:3b-preview-fp16":{"mode":"chat","base_model":"smallthinker:3b-preview-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smallthinker:3b-preview-q4_K_M":{"mode":"chat","base_model":"smallthinker:3b-preview-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"smallthinker:3b-preview-q8_0":{"mode":"chat","base_model":"smallthinker:3b-preview-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"paraphrase-multilingual:278m":{"mode":"chat","base_model":"paraphrase-multilingual:278m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"paraphrase-multilingual:278m-mpnet-base-v2-fp16":{"mode":"chat","base_model":"paraphrase-multilingual:278m-mpnet-base-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-fp16":{"mode":"chat","base_model":"llama-pro:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q2_K":{"mode":"chat","base_model":"llama-pro:8b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q3_K_L":{"mode":"chat","base_model":"llama-pro:8b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q3_K_M":{"mode":"chat","base_model":"llama-pro:8b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q3_K_S":{"mode":"chat","base_model":"llama-pro:8b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q4_0":{"mode":"chat","base_model":"llama-pro:8b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q4_1":{"mode":"chat","base_model":"llama-pro:8b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q4_K_M":{"mode":"chat","base_model":"llama-pro:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q4_K_S":{"mode":"chat","base_model":"llama-pro:8b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q5_0":{"mode":"chat","base_model":"llama-pro:8b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q5_1":{"mode":"chat","base_model":"llama-pro:8b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q5_K_M":{"mode":"chat","base_model":"llama-pro:8b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q5_K_S":{"mode":"chat","base_model":"llama-pro:8b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q6_K":{"mode":"chat","base_model":"llama-pro:8b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-instruct-q8_0":{"mode":"chat","base_model":"llama-pro:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-fp16":{"mode":"chat","base_model":"llama-pro:8b-text-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q2_K":{"mode":"chat","base_model":"llama-pro:8b-text-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q3_K_L":{"mode":"chat","base_model":"llama-pro:8b-text-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q3_K_M":{"mode":"chat","base_model":"llama-pro:8b-text-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q3_K_S":{"mode":"chat","base_model":"llama-pro:8b-text-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q4_0":{"mode":"chat","base_model":"llama-pro:8b-text-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q4_1":{"mode":"chat","base_model":"llama-pro:8b-text-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q4_K_M":{"mode":"chat","base_model":"llama-pro:8b-text-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q4_K_S":{"mode":"chat","base_model":"llama-pro:8b-text-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q5_0":{"mode":"chat","base_model":"llama-pro:8b-text-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q5_1":{"mode":"chat","base_model":"llama-pro:8b-text-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q5_K_M":{"mode":"chat","base_model":"llama-pro:8b-text-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q5_K_S":{"mode":"chat","base_model":"llama-pro:8b-text-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q6_K":{"mode":"chat","base_model":"llama-pro:8b-text-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:8b-text-q8_0":{"mode":"chat","base_model":"llama-pro:8b-text-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:instruct":{"mode":"chat","base_model":"llama-pro:instruct","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-pro:text":{"mode":"chat","base_model":"llama-pro:text","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b":{"mode":"chat","base_model":"yarn-mistral:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k":{"mode":"chat","base_model":"yarn-mistral:7b-128k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-fp16":{"mode":"chat","base_model":"yarn-mistral:7b-128k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q2_K":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q3_K_L":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q3_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q3_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q4_0":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q4_1":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q4_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q4_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q5_0":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q5_1":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q5_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q5_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q6_K":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-128k-q8_0":{"mode":"chat","base_model":"yarn-mistral:7b-128k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k":{"mode":"chat","base_model":"yarn-mistral:7b-64k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q2_K":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q3_K_L":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q3_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q3_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q4_0":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q4_1":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q4_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q4_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q5_0":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q5_1":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q5_K_M":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q5_K_S":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q6_K":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"yarn-mistral:7b-64k-q8_0":{"mode":"chat","base_model":"yarn-mistral:7b-64k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b":{"mode":"chat","base_model":"aya-expanse:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b":{"mode":"chat","base_model":"aya-expanse:32b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-fp16":{"mode":"chat","base_model":"aya-expanse:32b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q2_K":{"mode":"chat","base_model":"aya-expanse:32b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q3_K_L":{"mode":"chat","base_model":"aya-expanse:32b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q3_K_M":{"mode":"chat","base_model":"aya-expanse:32b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q3_K_S":{"mode":"chat","base_model":"aya-expanse:32b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q4_0":{"mode":"chat","base_model":"aya-expanse:32b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q4_1":{"mode":"chat","base_model":"aya-expanse:32b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q4_K_M":{"mode":"chat","base_model":"aya-expanse:32b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q4_K_S":{"mode":"chat","base_model":"aya-expanse:32b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q5_0":{"mode":"chat","base_model":"aya-expanse:32b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q5_1":{"mode":"chat","base_model":"aya-expanse:32b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q5_K_M":{"mode":"chat","base_model":"aya-expanse:32b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q5_K_S":{"mode":"chat","base_model":"aya-expanse:32b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q6_K":{"mode":"chat","base_model":"aya-expanse:32b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:32b-q8_0":{"mode":"chat","base_model":"aya-expanse:32b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-fp16":{"mode":"chat","base_model":"aya-expanse:8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q2_K":{"mode":"chat","base_model":"aya-expanse:8b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q3_K_L":{"mode":"chat","base_model":"aya-expanse:8b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q3_K_M":{"mode":"chat","base_model":"aya-expanse:8b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q3_K_S":{"mode":"chat","base_model":"aya-expanse:8b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q4_0":{"mode":"chat","base_model":"aya-expanse:8b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q4_1":{"mode":"chat","base_model":"aya-expanse:8b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q4_K_M":{"mode":"chat","base_model":"aya-expanse:8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q4_K_S":{"mode":"chat","base_model":"aya-expanse:8b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q5_0":{"mode":"chat","base_model":"aya-expanse:8b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q5_1":{"mode":"chat","base_model":"aya-expanse:8b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q5_K_M":{"mode":"chat","base_model":"aya-expanse:8b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q5_K_S":{"mode":"chat","base_model":"aya-expanse:8b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q6_K":{"mode":"chat","base_model":"aya-expanse:8b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"aya-expanse:8b-q8_0":{"mode":"chat","base_model":"aya-expanse:8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b":{"mode":"chat","base_model":"granite3-moe:1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b":{"mode":"chat","base_model":"granite3-moe:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-fp16":{"mode":"chat","base_model":"granite3-moe:1b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q2_K":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q4_0":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q4_1":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q5_0":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q5_1":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q6_K":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:1b-instruct-q8_0":{"mode":"chat","base_model":"granite3-moe:1b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-fp16":{"mode":"chat","base_model":"granite3-moe:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q2_K":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q4_0":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q4_1":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q5_0":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q5_1":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q6_K":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-moe:3b-instruct-q8_0":{"mode":"chat","base_model":"granite3-moe:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b":{"mode":"chat","base_model":"nexusraven:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-fp16":{"mode":"chat","base_model":"nexusraven:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q2_K":{"mode":"chat","base_model":"nexusraven:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q3_K_L":{"mode":"chat","base_model":"nexusraven:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q3_K_M":{"mode":"chat","base_model":"nexusraven:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q3_K_S":{"mode":"chat","base_model":"nexusraven:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q4_0":{"mode":"chat","base_model":"nexusraven:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q4_1":{"mode":"chat","base_model":"nexusraven:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q4_K_M":{"mode":"chat","base_model":"nexusraven:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q4_K_S":{"mode":"chat","base_model":"nexusraven:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q5_0":{"mode":"chat","base_model":"nexusraven:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q5_1":{"mode":"chat","base_model":"nexusraven:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q5_K_M":{"mode":"chat","base_model":"nexusraven:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q5_K_S":{"mode":"chat","base_model":"nexusraven:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q6_K":{"mode":"chat","base_model":"nexusraven:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-q8_0":{"mode":"chat","base_model":"nexusraven:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-fp16":{"mode":"chat","base_model":"nexusraven:13b-v2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q2_K":{"mode":"chat","base_model":"nexusraven:13b-v2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q3_K_L":{"mode":"chat","base_model":"nexusraven:13b-v2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q3_K_M":{"mode":"chat","base_model":"nexusraven:13b-v2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q3_K_S":{"mode":"chat","base_model":"nexusraven:13b-v2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q4_0":{"mode":"chat","base_model":"nexusraven:13b-v2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q4_1":{"mode":"chat","base_model":"nexusraven:13b-v2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q4_K_M":{"mode":"chat","base_model":"nexusraven:13b-v2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q4_K_S":{"mode":"chat","base_model":"nexusraven:13b-v2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q5_0":{"mode":"chat","base_model":"nexusraven:13b-v2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q5_1":{"mode":"chat","base_model":"nexusraven:13b-v2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q5_K_M":{"mode":"chat","base_model":"nexusraven:13b-v2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q5_K_S":{"mode":"chat","base_model":"nexusraven:13b-v2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q6_K":{"mode":"chat","base_model":"nexusraven:13b-v2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nexusraven:13b-v2-q8_0":{"mode":"chat","base_model":"nexusraven:13b-v2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:1b":{"mode":"chat","base_model":"falcon3:1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:3b":{"mode":"chat","base_model":"falcon3:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:7b":{"mode":"chat","base_model":"falcon3:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:10b":{"mode":"chat","base_model":"falcon3:10b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:10b-instruct-fp16":{"mode":"chat","base_model":"falcon3:10b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:10b-instruct-q4_K_M":{"mode":"chat","base_model":"falcon3:10b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:10b-instruct-q8_0":{"mode":"chat","base_model":"falcon3:10b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:1b-instruct-fp16":{"mode":"chat","base_model":"falcon3:1b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:1b-instruct-q4_K_M":{"mode":"chat","base_model":"falcon3:1b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:1b-instruct-q8_0":{"mode":"chat","base_model":"falcon3:1b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:3b-instruct-fp16":{"mode":"chat","base_model":"falcon3:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:3b-instruct-q4_K_M":{"mode":"chat","base_model":"falcon3:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:3b-instruct-q8_0":{"mode":"chat","base_model":"falcon3:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:7b-instruct-fp16":{"mode":"chat","base_model":"falcon3:7b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:7b-instruct-q4_K_M":{"mode":"chat","base_model":"falcon3:7b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon3:7b-instruct-q8_0":{"mode":"chat","base_model":"falcon3:7b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b":{"mode":"chat","base_model":"codeup:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2":{"mode":"chat","base_model":"codeup:13b-llama2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat":{"mode":"chat","base_model":"codeup:13b-llama2-chat","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-fp16":{"mode":"chat","base_model":"codeup:13b-llama2-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q2_K":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q3_K_L":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q3_K_M":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q3_K_S":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q4_0":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q4_1":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q4_K_M":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q4_K_S":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q5_0":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q5_1":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q5_K_M":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q5_K_S":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q6_K":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codeup:13b-llama2-chat-q8_0":{"mode":"chat","base_model":"codeup:13b-llama2-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-fp16":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q2_K":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q3_K_L":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q3_K_M":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q3_K_S":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q4_0":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q4_1":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q4_K_M":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q4_K_S":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q5_0":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q5_1":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q5_K_M":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q5_K_S":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q6_K":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:8x7b-dpo-q8_0":{"mode":"chat","base_model":"nous-hermes2-mixtral:8x7b-dpo-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nous-hermes2-mixtral:dpo":{"mode":"chat","base_model":"nous-hermes2-mixtral:dpo","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b":{"mode":"chat","base_model":"everythinglm:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k":{"mode":"chat","base_model":"everythinglm:13b-16k","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-fp16":{"mode":"chat","base_model":"everythinglm:13b-16k-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q2_K":{"mode":"chat","base_model":"everythinglm:13b-16k-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q3_K_L":{"mode":"chat","base_model":"everythinglm:13b-16k-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q3_K_M":{"mode":"chat","base_model":"everythinglm:13b-16k-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q3_K_S":{"mode":"chat","base_model":"everythinglm:13b-16k-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q4_0":{"mode":"chat","base_model":"everythinglm:13b-16k-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q4_1":{"mode":"chat","base_model":"everythinglm:13b-16k-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q4_K_M":{"mode":"chat","base_model":"everythinglm:13b-16k-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q4_K_S":{"mode":"chat","base_model":"everythinglm:13b-16k-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q5_0":{"mode":"chat","base_model":"everythinglm:13b-16k-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q5_1":{"mode":"chat","base_model":"everythinglm:13b-16k-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q5_K_M":{"mode":"chat","base_model":"everythinglm:13b-16k-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q5_K_S":{"mode":"chat","base_model":"everythinglm:13b-16k-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q6_K":{"mode":"chat","base_model":"everythinglm:13b-16k-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"everythinglm:13b-16k-q8_0":{"mode":"chat","base_model":"everythinglm:13b-16k-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b":{"mode":"chat","base_model":"shieldgemma:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b":{"mode":"chat","base_model":"shieldgemma:9b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b":{"mode":"chat","base_model":"shieldgemma:27b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-fp16":{"mode":"chat","base_model":"shieldgemma:27b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q2_K":{"mode":"chat","base_model":"shieldgemma:27b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q3_K_L":{"mode":"chat","base_model":"shieldgemma:27b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q3_K_M":{"mode":"chat","base_model":"shieldgemma:27b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q3_K_S":{"mode":"chat","base_model":"shieldgemma:27b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q4_0":{"mode":"chat","base_model":"shieldgemma:27b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q4_1":{"mode":"chat","base_model":"shieldgemma:27b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q4_K_M":{"mode":"chat","base_model":"shieldgemma:27b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q4_K_S":{"mode":"chat","base_model":"shieldgemma:27b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q5_0":{"mode":"chat","base_model":"shieldgemma:27b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q5_1":{"mode":"chat","base_model":"shieldgemma:27b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q5_K_M":{"mode":"chat","base_model":"shieldgemma:27b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q5_K_S":{"mode":"chat","base_model":"shieldgemma:27b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q6_K":{"mode":"chat","base_model":"shieldgemma:27b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:27b-q8_0":{"mode":"chat","base_model":"shieldgemma:27b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-fp16":{"mode":"chat","base_model":"shieldgemma:2b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q2_K":{"mode":"chat","base_model":"shieldgemma:2b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q3_K_L":{"mode":"chat","base_model":"shieldgemma:2b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q3_K_M":{"mode":"chat","base_model":"shieldgemma:2b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q3_K_S":{"mode":"chat","base_model":"shieldgemma:2b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q4_0":{"mode":"chat","base_model":"shieldgemma:2b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q4_1":{"mode":"chat","base_model":"shieldgemma:2b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q4_K_M":{"mode":"chat","base_model":"shieldgemma:2b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q4_K_S":{"mode":"chat","base_model":"shieldgemma:2b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q5_0":{"mode":"chat","base_model":"shieldgemma:2b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q5_1":{"mode":"chat","base_model":"shieldgemma:2b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q5_K_M":{"mode":"chat","base_model":"shieldgemma:2b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q5_K_S":{"mode":"chat","base_model":"shieldgemma:2b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q6_K":{"mode":"chat","base_model":"shieldgemma:2b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:2b-q8_0":{"mode":"chat","base_model":"shieldgemma:2b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-fp16":{"mode":"chat","base_model":"shieldgemma:9b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q2_K":{"mode":"chat","base_model":"shieldgemma:9b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q3_K_L":{"mode":"chat","base_model":"shieldgemma:9b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q3_K_M":{"mode":"chat","base_model":"shieldgemma:9b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q3_K_S":{"mode":"chat","base_model":"shieldgemma:9b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q4_0":{"mode":"chat","base_model":"shieldgemma:9b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q4_1":{"mode":"chat","base_model":"shieldgemma:9b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q4_K_M":{"mode":"chat","base_model":"shieldgemma:9b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q4_K_S":{"mode":"chat","base_model":"shieldgemma:9b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q5_0":{"mode":"chat","base_model":"shieldgemma:9b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q5_1":{"mode":"chat","base_model":"shieldgemma:9b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q5_K_M":{"mode":"chat","base_model":"shieldgemma:9b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q5_K_S":{"mode":"chat","base_model":"shieldgemma:9b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q6_K":{"mode":"chat","base_model":"shieldgemma:9b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"shieldgemma:9b-q8_0":{"mode":"chat","base_model":"shieldgemma:9b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b":{"mode":"chat","base_model":"granite3.1-moe:1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b":{"mode":"chat","base_model":"granite3.1-moe:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-fp16":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q2_K":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q4_0":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q4_1":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q5_0":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q5_1":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q6_K":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:1b-instruct-q8_0":{"mode":"chat","base_model":"granite3.1-moe:1b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-fp16":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q2_K":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q3_K_L":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q3_K_M":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q3_K_S":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q4_0":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q4_1":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q4_K_S":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q5_0":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q5_1":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q5_K_M":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q5_K_S":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q6_K":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.1-moe:3b-instruct-q8_0":{"mode":"chat","base_model":"granite3.1-moe:3b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed2:568m":{"mode":"chat","base_model":"snowflake-arctic-embed2:568m","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"snowflake-arctic-embed2:568m-l-fp16":{"mode":"chat","base_model":"snowflake-arctic-embed2:568m-l-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"marco-o1:7b":{"mode":"chat","base_model":"marco-o1:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"marco-o1:7b-fp16":{"mode":"chat","base_model":"marco-o1:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"marco-o1:7b-q4_K_M":{"mode":"chat","base_model":"marco-o1:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"marco-o1:7b-q8_0":{"mode":"chat","base_model":"marco-o1:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b":{"mode":"chat","base_model":"falcon2:11b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-fp16":{"mode":"chat","base_model":"falcon2:11b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q2_K":{"mode":"chat","base_model":"falcon2:11b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q3_K_L":{"mode":"chat","base_model":"falcon2:11b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q3_K_M":{"mode":"chat","base_model":"falcon2:11b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q3_K_S":{"mode":"chat","base_model":"falcon2:11b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q4_0":{"mode":"chat","base_model":"falcon2:11b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q4_1":{"mode":"chat","base_model":"falcon2:11b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q4_K_M":{"mode":"chat","base_model":"falcon2:11b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q4_K_S":{"mode":"chat","base_model":"falcon2:11b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q5_0":{"mode":"chat","base_model":"falcon2:11b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q5_1":{"mode":"chat","base_model":"falcon2:11b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q5_K_M":{"mode":"chat","base_model":"falcon2:11b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q5_K_S":{"mode":"chat","base_model":"falcon2:11b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q6_K":{"mode":"chat","base_model":"falcon2:11b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"falcon2:11b-q8_0":{"mode":"chat","base_model":"falcon2:11b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b":{"mode":"chat","base_model":"mathstral:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-fp16":{"mode":"chat","base_model":"mathstral:7b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q2_K":{"mode":"chat","base_model":"mathstral:7b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q3_K_L":{"mode":"chat","base_model":"mathstral:7b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q3_K_M":{"mode":"chat","base_model":"mathstral:7b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q3_K_S":{"mode":"chat","base_model":"mathstral:7b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q4_0":{"mode":"chat","base_model":"mathstral:7b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q4_1":{"mode":"chat","base_model":"mathstral:7b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q4_K_M":{"mode":"chat","base_model":"mathstral:7b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q4_K_S":{"mode":"chat","base_model":"mathstral:7b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q5_0":{"mode":"chat","base_model":"mathstral:7b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q5_1":{"mode":"chat","base_model":"mathstral:7b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q5_K_M":{"mode":"chat","base_model":"mathstral:7b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q5_K_S":{"mode":"chat","base_model":"mathstral:7b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q6_K":{"mode":"chat","base_model":"mathstral:7b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mathstral:7b-v0.1-q8_0":{"mode":"chat","base_model":"mathstral:7b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b":{"mode":"chat","base_model":"magicoder:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl":{"mode":"chat","base_model":"magicoder:7b-s-cl","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-fp16":{"mode":"chat","base_model":"magicoder:7b-s-cl-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q2_K":{"mode":"chat","base_model":"magicoder:7b-s-cl-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q3_K_L":{"mode":"chat","base_model":"magicoder:7b-s-cl-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q3_K_M":{"mode":"chat","base_model":"magicoder:7b-s-cl-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q3_K_S":{"mode":"chat","base_model":"magicoder:7b-s-cl-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q4_0":{"mode":"chat","base_model":"magicoder:7b-s-cl-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q4_1":{"mode":"chat","base_model":"magicoder:7b-s-cl-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q4_K_M":{"mode":"chat","base_model":"magicoder:7b-s-cl-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q4_K_S":{"mode":"chat","base_model":"magicoder:7b-s-cl-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q5_0":{"mode":"chat","base_model":"magicoder:7b-s-cl-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q5_1":{"mode":"chat","base_model":"magicoder:7b-s-cl-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q5_K_M":{"mode":"chat","base_model":"magicoder:7b-s-cl-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q5_K_S":{"mode":"chat","base_model":"magicoder:7b-s-cl-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q6_K":{"mode":"chat","base_model":"magicoder:7b-s-cl-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"magicoder:7b-s-cl-q8_0":{"mode":"chat","base_model":"magicoder:7b-s-cl-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b":{"mode":"chat","base_model":"stablelm-zephyr:3b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-fp16":{"mode":"chat","base_model":"stablelm-zephyr:3b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q2_K":{"mode":"chat","base_model":"stablelm-zephyr:3b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q3_K_L":{"mode":"chat","base_model":"stablelm-zephyr:3b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q3_K_M":{"mode":"chat","base_model":"stablelm-zephyr:3b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q3_K_S":{"mode":"chat","base_model":"stablelm-zephyr:3b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q4_0":{"mode":"chat","base_model":"stablelm-zephyr:3b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q4_1":{"mode":"chat","base_model":"stablelm-zephyr:3b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q4_K_M":{"mode":"chat","base_model":"stablelm-zephyr:3b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q4_K_S":{"mode":"chat","base_model":"stablelm-zephyr:3b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q5_0":{"mode":"chat","base_model":"stablelm-zephyr:3b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q5_1":{"mode":"chat","base_model":"stablelm-zephyr:3b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q5_K_M":{"mode":"chat","base_model":"stablelm-zephyr:3b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q5_K_S":{"mode":"chat","base_model":"stablelm-zephyr:3b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q6_K":{"mode":"chat","base_model":"stablelm-zephyr:3b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"stablelm-zephyr:3b-q8_0":{"mode":"chat","base_model":"stablelm-zephyr:3b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b":{"mode":"chat","base_model":"reader-lm:0.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b":{"mode":"chat","base_model":"reader-lm:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-fp16":{"mode":"chat","base_model":"reader-lm:0.5b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q2_K":{"mode":"chat","base_model":"reader-lm:0.5b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q3_K_L":{"mode":"chat","base_model":"reader-lm:0.5b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q3_K_M":{"mode":"chat","base_model":"reader-lm:0.5b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q3_K_S":{"mode":"chat","base_model":"reader-lm:0.5b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q4_0":{"mode":"chat","base_model":"reader-lm:0.5b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q4_1":{"mode":"chat","base_model":"reader-lm:0.5b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q4_K_M":{"mode":"chat","base_model":"reader-lm:0.5b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q4_K_S":{"mode":"chat","base_model":"reader-lm:0.5b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q5_0":{"mode":"chat","base_model":"reader-lm:0.5b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q5_1":{"mode":"chat","base_model":"reader-lm:0.5b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q5_K_M":{"mode":"chat","base_model":"reader-lm:0.5b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q5_K_S":{"mode":"chat","base_model":"reader-lm:0.5b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q6_K":{"mode":"chat","base_model":"reader-lm:0.5b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:0.5b-q8_0":{"mode":"chat","base_model":"reader-lm:0.5b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-fp16":{"mode":"chat","base_model":"reader-lm:1.5b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q2_K":{"mode":"chat","base_model":"reader-lm:1.5b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q3_K_L":{"mode":"chat","base_model":"reader-lm:1.5b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q3_K_M":{"mode":"chat","base_model":"reader-lm:1.5b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q3_K_S":{"mode":"chat","base_model":"reader-lm:1.5b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q4_0":{"mode":"chat","base_model":"reader-lm:1.5b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q4_1":{"mode":"chat","base_model":"reader-lm:1.5b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q4_K_M":{"mode":"chat","base_model":"reader-lm:1.5b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q4_K_S":{"mode":"chat","base_model":"reader-lm:1.5b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q5_0":{"mode":"chat","base_model":"reader-lm:1.5b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q5_1":{"mode":"chat","base_model":"reader-lm:1.5b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q5_K_M":{"mode":"chat","base_model":"reader-lm:1.5b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q5_K_S":{"mode":"chat","base_model":"reader-lm:1.5b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q6_K":{"mode":"chat","base_model":"reader-lm:1.5b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"reader-lm:1.5b-q8_0":{"mode":"chat","base_model":"reader-lm:1.5b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b":{"mode":"chat","base_model":"solar-pro:22b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-fp16":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q2_K":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q3_K_L":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q3_K_M":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q3_K_S":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q4_0":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q4_1":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q4_K_M":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q4_K_S":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q5_0":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q5_1":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q5_K_M":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q5_K_S":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q6_K":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:22b-preview-instruct-q8_0":{"mode":"chat","base_model":"solar-pro:22b-preview-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"solar-pro:preview":{"mode":"chat","base_model":"solar-pro:preview","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b":{"mode":"chat","base_model":"codebooga:34b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-fp16":{"mode":"chat","base_model":"codebooga:34b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q2_K":{"mode":"chat","base_model":"codebooga:34b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q3_K_L":{"mode":"chat","base_model":"codebooga:34b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q3_K_M":{"mode":"chat","base_model":"codebooga:34b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q3_K_S":{"mode":"chat","base_model":"codebooga:34b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q4_0":{"mode":"chat","base_model":"codebooga:34b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q4_1":{"mode":"chat","base_model":"codebooga:34b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q4_K_M":{"mode":"chat","base_model":"codebooga:34b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q5_0":{"mode":"chat","base_model":"codebooga:34b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q5_1":{"mode":"chat","base_model":"codebooga:34b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q5_K_M":{"mode":"chat","base_model":"codebooga:34b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q5_K_S":{"mode":"chat","base_model":"codebooga:34b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q6_K":{"mode":"chat","base_model":"codebooga:34b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"codebooga:34b-v0.1-q8_0":{"mode":"chat","base_model":"codebooga:34b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b":{"mode":"chat","base_model":"duckdb-nsql:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-fp16":{"mode":"chat","base_model":"duckdb-nsql:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q2_K":{"mode":"chat","base_model":"duckdb-nsql:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q3_K_L":{"mode":"chat","base_model":"duckdb-nsql:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q3_K_M":{"mode":"chat","base_model":"duckdb-nsql:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q3_K_S":{"mode":"chat","base_model":"duckdb-nsql:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q4_0":{"mode":"chat","base_model":"duckdb-nsql:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q4_1":{"mode":"chat","base_model":"duckdb-nsql:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q4_K_M":{"mode":"chat","base_model":"duckdb-nsql:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q4_K_S":{"mode":"chat","base_model":"duckdb-nsql:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q5_0":{"mode":"chat","base_model":"duckdb-nsql:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q5_1":{"mode":"chat","base_model":"duckdb-nsql:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q5_K_M":{"mode":"chat","base_model":"duckdb-nsql:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q5_K_S":{"mode":"chat","base_model":"duckdb-nsql:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q6_K":{"mode":"chat","base_model":"duckdb-nsql:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"duckdb-nsql:7b-q8_0":{"mode":"chat","base_model":"duckdb-nsql:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b":{"mode":"chat","base_model":"mistrallite:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-fp16":{"mode":"chat","base_model":"mistrallite:7b-v0.1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q2_K":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q3_K_L":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q3_K_M":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q3_K_S":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q4_0":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q4_1":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q4_K_M":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q4_K_S":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q5_0":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q5_1":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q5_K_M":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q5_K_S":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q6_K":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistrallite:7b-v0.1-q8_0":{"mode":"chat","base_model":"mistrallite:7b-v0.1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-fp16":{"mode":"chat","base_model":"llama-guard3:1b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q2_K":{"mode":"chat","base_model":"llama-guard3:1b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q3_K_L":{"mode":"chat","base_model":"llama-guard3:1b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q3_K_M":{"mode":"chat","base_model":"llama-guard3:1b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q3_K_S":{"mode":"chat","base_model":"llama-guard3:1b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q4_0":{"mode":"chat","base_model":"llama-guard3:1b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q4_1":{"mode":"chat","base_model":"llama-guard3:1b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q4_K_M":{"mode":"chat","base_model":"llama-guard3:1b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q4_K_S":{"mode":"chat","base_model":"llama-guard3:1b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q5_0":{"mode":"chat","base_model":"llama-guard3:1b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q5_1":{"mode":"chat","base_model":"llama-guard3:1b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q5_K_M":{"mode":"chat","base_model":"llama-guard3:1b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q5_K_S":{"mode":"chat","base_model":"llama-guard3:1b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q6_K":{"mode":"chat","base_model":"llama-guard3:1b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:1b-q8_0":{"mode":"chat","base_model":"llama-guard3:1b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-fp16":{"mode":"chat","base_model":"llama-guard3:8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q2_K":{"mode":"chat","base_model":"llama-guard3:8b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q3_K_L":{"mode":"chat","base_model":"llama-guard3:8b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q3_K_M":{"mode":"chat","base_model":"llama-guard3:8b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q3_K_S":{"mode":"chat","base_model":"llama-guard3:8b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q4_0":{"mode":"chat","base_model":"llama-guard3:8b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q4_1":{"mode":"chat","base_model":"llama-guard3:8b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q4_K_M":{"mode":"chat","base_model":"llama-guard3:8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q4_K_S":{"mode":"chat","base_model":"llama-guard3:8b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q5_0":{"mode":"chat","base_model":"llama-guard3:8b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q5_1":{"mode":"chat","base_model":"llama-guard3:8b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q5_K_M":{"mode":"chat","base_model":"llama-guard3:8b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q5_K_S":{"mode":"chat","base_model":"llama-guard3:8b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q6_K":{"mode":"chat","base_model":"llama-guard3:8b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"llama-guard3:8b-q8_0":{"mode":"chat","base_model":"llama-guard3:8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b":{"mode":"chat","base_model":"wizard-vicuna:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-fp16":{"mode":"chat","base_model":"wizard-vicuna:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q2_K":{"mode":"chat","base_model":"wizard-vicuna:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q3_K_L":{"mode":"chat","base_model":"wizard-vicuna:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q3_K_M":{"mode":"chat","base_model":"wizard-vicuna:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q3_K_S":{"mode":"chat","base_model":"wizard-vicuna:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q4_0":{"mode":"chat","base_model":"wizard-vicuna:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q4_1":{"mode":"chat","base_model":"wizard-vicuna:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q4_K_M":{"mode":"chat","base_model":"wizard-vicuna:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q4_K_S":{"mode":"chat","base_model":"wizard-vicuna:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q5_0":{"mode":"chat","base_model":"wizard-vicuna:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q5_1":{"mode":"chat","base_model":"wizard-vicuna:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q5_K_M":{"mode":"chat","base_model":"wizard-vicuna:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q5_K_S":{"mode":"chat","base_model":"wizard-vicuna:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q6_K":{"mode":"chat","base_model":"wizard-vicuna:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"wizard-vicuna:13b-q8_0":{"mode":"chat","base_model":"wizard-vicuna:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:2.4b":{"mode":"chat","base_model":"exaone3.5:2.4b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:7.8b":{"mode":"chat","base_model":"exaone3.5:7.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:32b":{"mode":"chat","base_model":"exaone3.5:32b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:2.4b-instruct-fp16":{"mode":"chat","base_model":"exaone3.5:2.4b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:2.4b-instruct-q4_K_M":{"mode":"chat","base_model":"exaone3.5:2.4b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:2.4b-instruct-q8_0":{"mode":"chat","base_model":"exaone3.5:2.4b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:32b-instruct-fp16":{"mode":"chat","base_model":"exaone3.5:32b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:32b-instruct-q4_K_M":{"mode":"chat","base_model":"exaone3.5:32b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:32b-instruct-q8_0":{"mode":"chat","base_model":"exaone3.5:32b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:7.8b-instruct-fp16":{"mode":"chat","base_model":"exaone3.5:7.8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:7.8b-instruct-q4_K_M":{"mode":"chat","base_model":"exaone3.5:7.8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"exaone3.5:7.8b-instruct-q8_0":{"mode":"chat","base_model":"exaone3.5:7.8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b":{"mode":"chat","base_model":"megadolphin:120b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2":{"mode":"chat","base_model":"megadolphin:120b-v2.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-fp16":{"mode":"chat","base_model":"megadolphin:120b-v2.2-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q2_K":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q3_K_L":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q3_K_M":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q3_K_S":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q4_0":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q4_1":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q4_K_M":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q4_K_S":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q5_0":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q5_1":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q5_K_M":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q5_K_S":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q6_K":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:120b-v2.2-q8_0":{"mode":"chat","base_model":"megadolphin:120b-v2.2-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"megadolphin:v2.2":{"mode":"chat","base_model":"megadolphin:v2.2","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b":{"mode":"chat","base_model":"nuextract:3.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-fp16":{"mode":"chat","base_model":"nuextract:3.8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q2_K":{"mode":"chat","base_model":"nuextract:3.8b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q3_K_L":{"mode":"chat","base_model":"nuextract:3.8b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q3_K_M":{"mode":"chat","base_model":"nuextract:3.8b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q3_K_S":{"mode":"chat","base_model":"nuextract:3.8b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q4_0":{"mode":"chat","base_model":"nuextract:3.8b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q4_1":{"mode":"chat","base_model":"nuextract:3.8b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q4_K_M":{"mode":"chat","base_model":"nuextract:3.8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q4_K_S":{"mode":"chat","base_model":"nuextract:3.8b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q5_0":{"mode":"chat","base_model":"nuextract:3.8b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q5_1":{"mode":"chat","base_model":"nuextract:3.8b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q5_K_M":{"mode":"chat","base_model":"nuextract:3.8b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q5_K_S":{"mode":"chat","base_model":"nuextract:3.8b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q6_K":{"mode":"chat","base_model":"nuextract:3.8b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"nuextract:3.8b-q8_0":{"mode":"chat","base_model":"nuextract:3.8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:1.5b":{"mode":"chat","base_model":"opencoder:1.5b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:8b":{"mode":"chat","base_model":"opencoder:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:1.5b-instruct-fp16":{"mode":"chat","base_model":"opencoder:1.5b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:1.5b-instruct-q4_K_M":{"mode":"chat","base_model":"opencoder:1.5b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:1.5b-instruct-q8_0":{"mode":"chat","base_model":"opencoder:1.5b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:8b-instruct-fp16":{"mode":"chat","base_model":"opencoder:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:8b-instruct-q4_K_M":{"mode":"chat","base_model":"opencoder:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"opencoder:8b-instruct-q8_0":{"mode":"chat","base_model":"opencoder:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b":{"mode":"chat","base_model":"notux:8x7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1":{"mode":"chat","base_model":"notux:8x7b-v1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-fp16":{"mode":"chat","base_model":"notux:8x7b-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q2_K":{"mode":"chat","base_model":"notux:8x7b-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q3_K_L":{"mode":"chat","base_model":"notux:8x7b-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q3_K_M":{"mode":"chat","base_model":"notux:8x7b-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q3_K_S":{"mode":"chat","base_model":"notux:8x7b-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q4_0":{"mode":"chat","base_model":"notux:8x7b-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q4_1":{"mode":"chat","base_model":"notux:8x7b-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q4_K_M":{"mode":"chat","base_model":"notux:8x7b-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q4_K_S":{"mode":"chat","base_model":"notux:8x7b-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q5_0":{"mode":"chat","base_model":"notux:8x7b-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q5_1":{"mode":"chat","base_model":"notux:8x7b-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q5_K_M":{"mode":"chat","base_model":"notux:8x7b-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q5_K_S":{"mode":"chat","base_model":"notux:8x7b-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q6_K":{"mode":"chat","base_model":"notux:8x7b-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notux:8x7b-v1-q8_0":{"mode":"chat","base_model":"notux:8x7b-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b":{"mode":"chat","base_model":"open-orca-platypus2:13b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-fp16":{"mode":"chat","base_model":"open-orca-platypus2:13b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q2_K":{"mode":"chat","base_model":"open-orca-platypus2:13b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q3_K_L":{"mode":"chat","base_model":"open-orca-platypus2:13b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q3_K_M":{"mode":"chat","base_model":"open-orca-platypus2:13b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q3_K_S":{"mode":"chat","base_model":"open-orca-platypus2:13b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q4_0":{"mode":"chat","base_model":"open-orca-platypus2:13b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q4_1":{"mode":"chat","base_model":"open-orca-platypus2:13b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q4_K_M":{"mode":"chat","base_model":"open-orca-platypus2:13b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q4_K_S":{"mode":"chat","base_model":"open-orca-platypus2:13b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q5_0":{"mode":"chat","base_model":"open-orca-platypus2:13b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q5_1":{"mode":"chat","base_model":"open-orca-platypus2:13b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q5_K_M":{"mode":"chat","base_model":"open-orca-platypus2:13b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q5_K_S":{"mode":"chat","base_model":"open-orca-platypus2:13b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q6_K":{"mode":"chat","base_model":"open-orca-platypus2:13b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"open-orca-platypus2:13b-q8_0":{"mode":"chat","base_model":"open-orca-platypus2:13b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b":{"mode":"chat","base_model":"notus:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1":{"mode":"chat","base_model":"notus:7b-v1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-fp16":{"mode":"chat","base_model":"notus:7b-v1-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q2_K":{"mode":"chat","base_model":"notus:7b-v1-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q3_K_L":{"mode":"chat","base_model":"notus:7b-v1-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q3_K_M":{"mode":"chat","base_model":"notus:7b-v1-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q3_K_S":{"mode":"chat","base_model":"notus:7b-v1-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q4_0":{"mode":"chat","base_model":"notus:7b-v1-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q4_1":{"mode":"chat","base_model":"notus:7b-v1-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q4_K_M":{"mode":"chat","base_model":"notus:7b-v1-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q4_K_S":{"mode":"chat","base_model":"notus:7b-v1-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q5_0":{"mode":"chat","base_model":"notus:7b-v1-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q5_1":{"mode":"chat","base_model":"notus:7b-v1-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q5_K_M":{"mode":"chat","base_model":"notus:7b-v1-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q5_K_S":{"mode":"chat","base_model":"notus:7b-v1-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q6_K":{"mode":"chat","base_model":"notus:7b-v1-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"notus:7b-v1-q8_0":{"mode":"chat","base_model":"notus:7b-v1-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-fp16":{"mode":"chat","base_model":"goliath:120b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q2_K":{"mode":"chat","base_model":"goliath:120b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q3_K_L":{"mode":"chat","base_model":"goliath:120b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q3_K_M":{"mode":"chat","base_model":"goliath:120b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q3_K_S":{"mode":"chat","base_model":"goliath:120b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q4_0":{"mode":"chat","base_model":"goliath:120b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q4_1":{"mode":"chat","base_model":"goliath:120b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q4_K_M":{"mode":"chat","base_model":"goliath:120b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q4_K_S":{"mode":"chat","base_model":"goliath:120b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q5_0":{"mode":"chat","base_model":"goliath:120b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q5_1":{"mode":"chat","base_model":"goliath:120b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q5_K_M":{"mode":"chat","base_model":"goliath:120b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q5_K_S":{"mode":"chat","base_model":"goliath:120b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q6_K":{"mode":"chat","base_model":"goliath:120b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"goliath:120b-q8_0":{"mode":"chat","base_model":"goliath:120b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b":{"mode":"chat","base_model":"bespoke-minicheck:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-fp16":{"mode":"chat","base_model":"bespoke-minicheck:7b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q2_K":{"mode":"chat","base_model":"bespoke-minicheck:7b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q3_K_L":{"mode":"chat","base_model":"bespoke-minicheck:7b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q3_K_M":{"mode":"chat","base_model":"bespoke-minicheck:7b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q3_K_S":{"mode":"chat","base_model":"bespoke-minicheck:7b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q4_0":{"mode":"chat","base_model":"bespoke-minicheck:7b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q4_1":{"mode":"chat","base_model":"bespoke-minicheck:7b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q4_K_M":{"mode":"chat","base_model":"bespoke-minicheck:7b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q4_K_S":{"mode":"chat","base_model":"bespoke-minicheck:7b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q5_0":{"mode":"chat","base_model":"bespoke-minicheck:7b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q5_1":{"mode":"chat","base_model":"bespoke-minicheck:7b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q5_K_M":{"mode":"chat","base_model":"bespoke-minicheck:7b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q5_K_S":{"mode":"chat","base_model":"bespoke-minicheck:7b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q6_K":{"mode":"chat","base_model":"bespoke-minicheck:7b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bespoke-minicheck:7b-q8_0":{"mode":"chat","base_model":"bespoke-minicheck:7b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r7b:7b":{"mode":"chat","base_model":"command-r7b:7b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r7b:7b-12-2024-fp16":{"mode":"chat","base_model":"command-r7b:7b-12-2024-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r7b:7b-12-2024-q4_K_M":{"mode":"chat","base_model":"command-r7b:7b-12-2024-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"command-r7b:7b-12-2024-q8_0":{"mode":"chat","base_model":"command-r7b:7b-12-2024-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b":{"mode":"chat","base_model":"firefunction-v2:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-fp16":{"mode":"chat","base_model":"firefunction-v2:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q2_K":{"mode":"chat","base_model":"firefunction-v2:70b-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q3_K_L":{"mode":"chat","base_model":"firefunction-v2:70b-q3_K_L","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q3_K_M":{"mode":"chat","base_model":"firefunction-v2:70b-q3_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q3_K_S":{"mode":"chat","base_model":"firefunction-v2:70b-q3_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q4_0":{"mode":"chat","base_model":"firefunction-v2:70b-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q4_1":{"mode":"chat","base_model":"firefunction-v2:70b-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q4_K_M":{"mode":"chat","base_model":"firefunction-v2:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q4_K_S":{"mode":"chat","base_model":"firefunction-v2:70b-q4_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q5_0":{"mode":"chat","base_model":"firefunction-v2:70b-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q5_1":{"mode":"chat","base_model":"firefunction-v2:70b-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q5_K_M":{"mode":"chat","base_model":"firefunction-v2:70b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q5_K_S":{"mode":"chat","base_model":"firefunction-v2:70b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q6_K":{"mode":"chat","base_model":"firefunction-v2:70b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"firefunction-v2:70b-q8_0":{"mode":"chat","base_model":"firefunction-v2:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:8b":{"mode":"chat","base_model":"tulu3:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:70b":{"mode":"chat","base_model":"tulu3:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:70b-fp16":{"mode":"chat","base_model":"tulu3:70b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:70b-q4_K_M":{"mode":"chat","base_model":"tulu3:70b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:70b-q8_0":{"mode":"chat","base_model":"tulu3:70b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:8b-fp16":{"mode":"chat","base_model":"tulu3:8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:8b-q4_K_M":{"mode":"chat","base_model":"tulu3:8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"tulu3:8b-q8_0":{"mode":"chat","base_model":"tulu3:8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dbrx:132b":{"mode":"chat","base_model":"dbrx:132b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dbrx:132b-instruct-fp16":{"mode":"chat","base_model":"dbrx:132b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dbrx:132b-instruct-q2_K":{"mode":"chat","base_model":"dbrx:132b-instruct-q2_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dbrx:132b-instruct-q4_0":{"mode":"chat","base_model":"dbrx:132b-instruct-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"dbrx:132b-instruct-q8_0":{"mode":"chat","base_model":"dbrx:132b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-embedding:30m":{"mode":"embedding","base_model":"granite-embedding:30m","provider":"ollama","max_input_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-embedding:278m":{"mode":"embedding","base_model":"granite-embedding:278m","provider":"ollama","max_input_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-embedding:278m-fp16":{"mode":"embedding","base_model":"granite-embedding:278m-fp16","provider":"ollama","max_input_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-embedding:30m-en":{"mode":"embedding","base_model":"granite-embedding:30m-en","provider":"ollama","max_input_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite-embedding:30m-en-fp16":{"mode":"embedding","base_model":"granite-embedding:30m-en-fp16","provider":"ollama","max_input_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:2b":{"mode":"chat","base_model":"granite3-guardian:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b":{"mode":"chat","base_model":"granite3-guardian:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:2b-fp16":{"mode":"chat","base_model":"granite3-guardian:2b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:2b-q8_0":{"mode":"chat","base_model":"granite3-guardian:2b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b-fp16":{"mode":"chat","base_model":"granite3-guardian:8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b-q5_K_M":{"mode":"chat","base_model":"granite3-guardian:8b-q5_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b-q5_K_S":{"mode":"chat","base_model":"granite3-guardian:8b-q5_K_S","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b-q6_K":{"mode":"chat","base_model":"granite3-guardian:8b-q6_K","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3-guardian:8b-q8_0":{"mode":"chat","base_model":"granite3-guardian:8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b":{"mode":"chat","base_model":"alfred:40b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b-1023-q4_0":{"mode":"chat","base_model":"alfred:40b-1023-q4_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b-1023-q4_1":{"mode":"chat","base_model":"alfred:40b-1023-q4_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b-1023-q5_0":{"mode":"chat","base_model":"alfred:40b-1023-q5_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b-1023-q5_1":{"mode":"chat","base_model":"alfred:40b-1023-q5_1","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"alfred:40b-1023-q8_0":{"mode":"chat","base_model":"alfred:40b-1023-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:1b":{"mode":"chat","base_model":"sailor2:1b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:8b":{"mode":"chat","base_model":"sailor2:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:20b":{"mode":"chat","base_model":"sailor2:20b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:1b-chat-fp16":{"mode":"chat","base_model":"sailor2:1b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:1b-chat-q4_K_M":{"mode":"chat","base_model":"sailor2:1b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:1b-chat-q8_0":{"mode":"chat","base_model":"sailor2:1b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:20b-chat-fp16":{"mode":"chat","base_model":"sailor2:20b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:20b-chat-q4_K_M":{"mode":"chat","base_model":"sailor2:20b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:20b-chat-q8_0":{"mode":"chat","base_model":"sailor2:20b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:8b-chat-fp16":{"mode":"chat","base_model":"sailor2:8b-chat-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:8b-chat-q4_K_M":{"mode":"chat","base_model":"sailor2:8b-chat-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"sailor2:8b-chat-q8_0":{"mode":"chat","base_model":"sailor2:8b-chat-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:70b":{"mode":"chat","base_model":"r1-1776:70b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:671b":{"mode":"chat","base_model":"r1-1776:671b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:671b-fp16":{"mode":"chat","base_model":"r1-1776:671b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:671b-q4_K_M":{"mode":"chat","base_model":"r1-1776:671b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:671b-q8_0":{"mode":"chat","base_model":"r1-1776:671b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:70b-distill-llama-fp16":{"mode":"chat","base_model":"r1-1776:70b-distill-llama-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:70b-distill-llama-q4_K_M":{"mode":"chat","base_model":"r1-1776:70b-distill-llama-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"r1-1776:70b-distill-llama-q8_0":{"mode":"chat","base_model":"r1-1776:70b-distill-llama-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:2b":{"mode":"chat","base_model":"granite3.2:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:8b":{"mode":"chat","base_model":"granite3.2:8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:2b-instruct-fp16":{"mode":"chat","base_model":"granite3.2:2b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:2b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.2:2b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:2b-instruct-q8_0":{"mode":"chat","base_model":"granite3.2:2b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:8b-instruct-fp16":{"mode":"chat","base_model":"granite3.2:8b-instruct-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:8b-instruct-q4_K_M":{"mode":"chat","base_model":"granite3.2:8b-instruct-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2:8b-instruct-q8_0":{"mode":"chat","base_model":"granite3.2:8b-instruct-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"granite3.2-vision:2b":{"mode":"chat","base_model":"granite3.2-vision:2b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"granite3.2-vision:2b-fp16":{"mode":"chat","base_model":"granite3.2-vision:2b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"granite3.2-vision:2b-q4_K_M":{"mode":"chat","base_model":"granite3.2-vision:2b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"granite3.2-vision:2b-q8_0":{"mode":"chat","base_model":"granite3.2-vision:2b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"phi4-mini:3.8b":{"mode":"chat","base_model":"phi4-mini:3.8b","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4-mini:3.8b-fp16":{"mode":"chat","base_model":"phi4-mini:3.8b-fp16","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4-mini:3.8b-q4_K_M":{"mode":"chat","base_model":"phi4-mini:3.8b-q4_K_M","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"phi4-mini:3.8b-q8_0":{"mode":"chat","base_model":"phi4-mini:3.8b-q8_0","provider":"ollama","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-3.5-turbo-0125":{"mode":"chat","base_model":"openai/gpt-3.5-turbo-0125","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true},"openai/gpt-3.5-turbo-1106":{"mode":"chat","base_model":"openai/gpt-3.5-turbo-1106","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true},"openai/gpt-3.5-turbo":{"mode":"chat","base_model":"openai/gpt-3.5-turbo","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4":{"mode":"chat","base_model":"openai/gpt-4","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"openai/gpt-4-0613":{"mode":"chat","base_model":"openai/gpt-4-0613","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_prompt_caching":true,"supports_system_messages":true},"openai/gpt-4-0125-preview":{"mode":"chat","base_model":"openai/gpt-4-0125-preview","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true},"openai/gpt-4-vision-preview":{"mode":"image_generation","base_model":"openai/gpt-4-vision-preview","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4-1106-vision-preview":{"mode":"image_generation","base_model":"openai/gpt-4-1106-vision-preview","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4o":{"mode":"image_generation","base_model":"openai/gpt-4o","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_web_search":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/gpt-4o-2024-05-13":{"mode":"image_generation","base_model":"openai/gpt-4o-2024-05-13","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_pdf_input":true},"openai/o1-preview":{"mode":"chat","base_model":"openai/o1-preview","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openai/o1":{"mode":"image_generation","base_model":"openai/o1","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"openai/o1-preview-2024-09-12":{"mode":"chat","base_model":"openai/o1-preview-2024-09-12","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openai/o3-mini":{"mode":"chat","base_model":"openai/o3-mini","provider":"litellm","max_input_tokens":100000,"max_output_tokens":100000,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"openai/gpt-4.5-preview":{"mode":"image_generation","base_model":"openai/gpt-4.5-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4.5-preview-2025-02-27":{"mode":"image_generation","base_model":"openai/gpt-4.5-preview-2025-02-27","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4.1-2025-04-14":{"mode":"image_generation","base_model":"openai/gpt-4.1-2025-04-14","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true},"openai/gpt-4.1":{"mode":"image_generation","base_model":"openai/gpt-4.1","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/gpt-4.1-mini-2025-04-14":{"mode":"image_generation","base_model":"openai/gpt-4.1-mini-2025-04-14","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true},"openai/gpt-4.1-mini":{"mode":"image_generation","base_model":"openai/gpt-4.1-mini","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/gpt-4.1-nano-2025-04-14":{"mode":"image_generation","base_model":"openai/gpt-4.1-nano-2025-04-14","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true},"openai/gpt-4.1-nano":{"mode":"image_generation","base_model":"openai/gpt-4.1-nano","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openai/o3-2025-04-16":{"mode":"image_generation","base_model":"openai/o3-2025-04-16","provider":"litellm","max_input_tokens":100000,"max_output_tokens":100000,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"openai/o3-mini-2025-01-31":{"mode":"image_generation","base_model":"openai/o3-mini-2025-01-31","provider":"litellm","max_input_tokens":100000,"max_output_tokens":100000,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false},"openai/o4-mini-2025-04-16":{"mode":"chat","base_model":"openai/o4-mini-2025-04-16","provider":"litellm","max_input_tokens":100000,"max_output_tokens":100000,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","default":"medium","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"openai/o1-mini-2024-09-12":{"mode":"chat","base_model":"openai/o1-mini-2024-09-12","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false},"openai/o1-mini":{"mode":"chat","base_model":"openai/o1-mini","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true},"openai/o1-2024-12-17":{"mode":"image_generation","base_model":"openai/o1-2024-12-17","provider":"litellm","max_input_tokens":100000,"max_output_tokens":100000,"max_tokens":100000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":100000,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true},"openai/gpt-4o-audio-preview":{"mode":"chat","base_model":"openai/gpt-4o-audio-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4o-audio-preview-2025-06-03":{"mode":"chat","base_model":"openai/gpt-4o-audio-preview-2025-06-03","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"openai/gpt-4o-audio-preview-2024-12-17":{"mode":"chat","base_model":"openai/gpt-4o-audio-preview-2024-12-17","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"openai/gpt-4o-audio-preview-2024-10-01":{"mode":"chat","base_model":"openai/gpt-4o-audio-preview-2024-10-01","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4o-mini-audio-preview":{"mode":"chat","base_model":"openai/gpt-4o-mini-audio-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4o-realtime-preview-2025-06-03":{"mode":"chat","base_model":"openai/gpt-4o-realtime-preview-2025-06-03","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"instructions","label":"Instructions","helpText":"The default system instructions (i.e. system message) prepended to model calls.","type":"text","supported_endpoints":["/v1/responses","/v1/realtime"]},{"id":"voice","label":"Voice","helpText":"Pre-selected voice used when generating the audio","type":"select","default":"alloy","options":[{"label":"Alloy","value":"alloy"},{"label":"Ash","value":"ash"},{"label":"Ballad","value":"ballad"},{"label":"Coral","value":"coral"},{"label":"Echo","value":"echo"},{"label":"Sage","value":"sage"},{"label":"Shimmer","value":"shimmer"},{"label":"Verse","value":"verse"}],"supported_endpoints":["/v1/chat/completions","/v1/realtime","/v1/audio/speech"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0.6,"max":1.2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_response_output_tokens","label":"Max Response Output Tokens","helpText":"Maximum number of output tokens for a single assistant response, inclusive of tool calls.","type":"number","default":510,"supported_endpoints":["/v1/realtime"]},{"id":"input_audio_noise_reduction","label":"Input Audio Noise Reduction","helpText":"Noise reduction applied to audio input, helpful with VAD and model understanding.","type":"select","accesorKey":"type","options":[{"label":"None","value":"none"},{"label":"Near Field","value":"near_field"},{"label":"Far Field","value":"far_field"}],"supported_endpoints":["/v1/realtime"]},{"id":"speed","label":"Speed","helpText":"The speed of the model's spoken response. ","type":"number","default":1,"range":{"min":0.25,"max":1.5,"step":0.05},"supported_endpoints":["/v1/realtime","/v1/audio/speech"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-4o-realtime-preview":{"mode":"chat","base_model":"openai/gpt-4o-realtime-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"instructions","label":"Instructions","helpText":"The default system instructions (i.e. system message) prepended to model calls.","type":"text","supported_endpoints":["/v1/responses","/v1/realtime"]},{"id":"voice","label":"Voice","helpText":"Pre-selected voice used when generating the audio","type":"select","default":"alloy","options":[{"label":"Alloy","value":"alloy"},{"label":"Ash","value":"ash"},{"label":"Ballad","value":"ballad"},{"label":"Coral","value":"coral"},{"label":"Echo","value":"echo"},{"label":"Sage","value":"sage"},{"label":"Shimmer","value":"shimmer"},{"label":"Verse","value":"verse"}],"supported_endpoints":["/v1/chat/completions","/v1/realtime","/v1/audio/speech"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0.6,"max":1.2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_response_output_tokens","label":"Max Response Output Tokens","helpText":"Maximum number of output tokens for a single assistant response, inclusive of tool calls.","type":"number","default":510,"supported_endpoints":["/v1/realtime"]},{"id":"input_audio_noise_reduction","label":"Input Audio Noise Reduction","helpText":"Noise reduction applied to audio input, helpful with VAD and model understanding.","type":"select","accesorKey":"type","options":[{"label":"None","value":"none"},{"label":"Near Field","value":"near_field"},{"label":"Far Field","value":"far_field"}],"supported_endpoints":["/v1/realtime"]},{"id":"speed","label":"Speed","helpText":"The speed of the model's spoken response. ","type":"number","default":1,"range":{"min":0.25,"max":1.5,"step":0.05},"supported_endpoints":["/v1/realtime","/v1/audio/speech"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openai/gpt-5-2025-08-07":{"mode":"image_generation","base_model":"openai/gpt-5-2025-08-07","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"supports_custom_tools":true},"openai/gpt-5":{"mode":"image_generation","base_model":"openai/gpt-5","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true},"openai/gpt-5-mini":{"mode":"image_generation","base_model":"openai/gpt-5-mini","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true},"openai/gpt-5-mini-2025-08-07":{"mode":"image_generation","base_model":"openai/gpt-5-mini-2025-08-07","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"supports_custom_tools":true},"openai/gpt-5-nano":{"mode":"image_generation","base_model":"openai/gpt-5-nano","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openai/gpt-5-nano-2025-08-07":{"mode":"image_generation","base_model":"openai/gpt-5-nano-2025-08-07","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true,"supports_custom_tools":true},"openai/gpt-5-chat":{"mode":"image_generation","base_model":"openai/gpt-5-chat","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_vision":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_minimal_reasoning_effort":true},"openai/gpt-5-chat-latest":{"mode":"image_generation","base_model":"openai/gpt-5-chat-latest","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_vision":true},"openai/gpt-oss-120b":{"mode":"chat","base_model":"openai/gpt-oss-120b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_output_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"supported_endpoints":["/v1/responses"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_web_search":true},"together/openai/gpt-oss-120b":{"mode":"chat","base_model":"together/openai/gpt-oss-120b","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"together/zero-one-ai/Yi-34B-Chat":{"mode":"chat","base_model":"together/zero-one-ai/Yi-34B-Chat","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Presence Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort is used to give the model guidance on how many reasoning tokens it should generate before creating a response to the prompt. You can specify one of low, medium, or high for this parameter, where low will favor speed and economical token usage, and high will favor more complete reasoning at the cost of more tokens generated and slower responses. The default value is medium, which is a balance between speed and reasoning accuracy.","type":"select","accesorKey":"type","default":{"type":"medium"},"options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"together/Austism/chronos-hermes-13b":{"mode":"chat","base_model":"together/Austism/chronos-hermes-13b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/DiscoResearch/DiscoLM-mixtral-8x7b-v2":{"mode":"chat","base_model":"together/DiscoResearch/DiscoLM-mixtral-8x7b-v2","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Gryphe/MythoMax-L2-13b":{"mode":"chat","base_model":"together/Gryphe/MythoMax-L2-13b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_function_calling":true},"together/lmsys/vicuna-13b-v1.5":{"mode":"chat","base_model":"together/lmsys/vicuna-13b-v1.5","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/lmsys/vicuna-7b-v1.5":{"mode":"chat","base_model":"together/lmsys/vicuna-7b-v1.5","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/lmsys/vicuna-13b-v1.5-16k":{"mode":"chat","base_model":"together/lmsys/vicuna-13b-v1.5-16k","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-13b-Instruct-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-13b-Instruct-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-34b-Instruct-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-34b-Instruct-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-70b-Instruct-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-70b-Instruct-hf","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-7b-Instruct-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-7b-Instruct-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/llama-2-13b-chat":{"mode":"chat","base_model":"together/togethercomputer/llama-2-13b-chat","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_tool_choice":true},"together/togethercomputer/llama-2-70b-chat":{"mode":"chat","base_model":"together/togethercomputer/llama-2-70b-chat","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/llama-2-7b-chat":{"mode":"chat","base_model":"together/togethercomputer/llama-2-7b-chat","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_tool_choice":true},"together/NousResearch/Nous-Capybara-7B-V1p9":{"mode":"chat","base_model":"together/NousResearch/Nous-Capybara-7B-V1p9","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-Llama2-70b":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-Llama2-70b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-llama-2-7b":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-llama-2-7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-Llama2-13b":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-Llama2-13b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-2-Yi-34B":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-2-Yi-34B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/openchat/openchat-3.5-1210":{"mode":"chat","base_model":"together/openchat/openchat-3.5-1210","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Open-Orca/Mistral-7B-OpenOrca":{"mode":"chat","base_model":"together/Open-Orca/Mistral-7B-OpenOrca","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Qwen-7B-Chat":{"mode":"chat","base_model":"together/togethercomputer/Qwen-7B-Chat","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/snorkelai/Snorkel-Mistral-PairRM-DPO":{"mode":"chat","base_model":"together/snorkelai/Snorkel-Mistral-PairRM-DPO","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/alpaca-7b":{"mode":"chat","base_model":"together/togethercomputer/alpaca-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/falcon-40b-instruct":{"mode":"chat","base_model":"together/togethercomputer/falcon-40b-instruct","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/falcon-7b-instruct":{"mode":"chat","base_model":"together/togethercomputer/falcon-7b-instruct","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/GPT-NeoXT-Chat-Base-20B":{"mode":"chat","base_model":"together/togethercomputer/GPT-NeoXT-Chat-Base-20B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Llama-2-7B-32K-Instruct":{"mode":"chat","base_model":"together/togethercomputer/Llama-2-7B-32K-Instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Pythia-Chat-Base-7B-v0.16":{"mode":"chat","base_model":"together/togethercomputer/Pythia-Chat-Base-7B-v0.16","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-Chat-3B-v1":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-Chat-3B-v1","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-7B-Chat":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-7B-Chat","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/StripedHyena-Nous-7B":{"mode":"chat","base_model":"together/togethercomputer/StripedHyena-Nous-7B","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Undi95/ReMM-SLERP-L2-13B":{"mode":"chat","base_model":"together/Undi95/ReMM-SLERP-L2-13B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_response_schema":true},"together/Undi95/Toppy-M-7B":{"mode":"chat","base_model":"together/Undi95/Toppy-M-7B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/WizardLM/WizardLM-13B-V1.2":{"mode":"chat","base_model":"together/WizardLM/WizardLM-13B-V1.2","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/garage-bAInd/Platypus2-70B-instruct":{"mode":"chat","base_model":"together/garage-bAInd/Platypus2-70B-instruct","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mistralai/Mistral-7B-Instruct-v0.1":{"mode":"chat","base_model":"together/mistralai/Mistral-7B-Instruct-v0.1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","type":"select","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/mistralai/Mistral-7B-Instruct-v0.2":{"mode":"chat","base_model":"together/mistralai/Mistral-7B-Instruct-v0.2","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mistralai/Mixtral-8x7B-Instruct-v0.1":{"mode":"chat","base_model":"together/mistralai/Mixtral-8x7B-Instruct-v0.1","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","type":"select","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/teknium/OpenHermes-2-Mistral-7B":{"mode":"chat","base_model":"together/teknium/OpenHermes-2-Mistral-7B","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/teknium/OpenHermes-2p5-Mistral-7B":{"mode":"chat","base_model":"together/teknium/OpenHermes-2p5-Mistral-7B","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/upstage/SOLAR-10.7B-Instruct-v1.0":{"mode":"chat","base_model":"together/upstage/SOLAR-10.7B-Instruct-v1.0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/zero-one-ai/Yi-34B":{"mode":"chat","base_model":"together/zero-one-ai/Yi-34B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/zero-one-ai/Yi-6B":{"mode":"chat","base_model":"together/zero-one-ai/Yi-6B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/llemma_7b":{"mode":"chat","base_model":"together/EleutherAI/llemma_7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/huggyllama/llama-65b":{"mode":"chat","base_model":"together/huggyllama/llama-65b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/llama-2-13b":{"mode":"chat","base_model":"together/togethercomputer/llama-2-13b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"together/togethercomputer/llama-2-70b":{"mode":"chat","base_model":"together/togethercomputer/llama-2-70b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"together/togethercomputer/llama-2-7b":{"mode":"chat","base_model":"together/togethercomputer/llama-2-7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"together/microsoft/phi-2":{"mode":"chat","base_model":"together/microsoft/phi-2","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Nexusflow/NexusRaven-V2-13B":{"mode":"chat","base_model":"together/Nexusflow/NexusRaven-V2-13B","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Qwen-7B":{"mode":"chat","base_model":"together/togethercomputer/Qwen-7B","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/falcon-40b":{"mode":"chat","base_model":"together/togethercomputer/falcon-40b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/falcon-7b":{"mode":"chat","base_model":"together/togethercomputer/falcon-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/GPT-JT-6B-v1":{"mode":"chat","base_model":"together/togethercomputer/GPT-JT-6B-v1","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/GPT-JT-Moderation-6B":{"mode":"chat","base_model":"together/togethercomputer/GPT-JT-Moderation-6B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/LLaMA-2-7B-32K":{"mode":"chat","base_model":"together/togethercomputer/LLaMA-2-7B-32K","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-Base-3B-v1":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-Base-3B-v1","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-7B-Base":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-7B-Base","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-Instruct-3B-v1":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-Instruct-3B-v1","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/RedPajama-INCITE-7B-Instruct":{"mode":"chat","base_model":"together/togethercomputer/RedPajama-INCITE-7B-Instruct","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/StripedHyena-Hessian-7B":{"mode":"chat","base_model":"together/togethercomputer/StripedHyena-Hessian-7B","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/WizardLM/WizardLM-70B-V1.0":{"mode":"chat","base_model":"together/WizardLM/WizardLM-70B-V1.0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mistralai/Mistral-7B-v0.1":{"mode":"chat","base_model":"together/mistralai/Mistral-7B-v0.1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mistralai/Mixtral-8x7B-v0.1":{"mode":"chat","base_model":"together/mistralai/Mixtral-8x7B-v0.1","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-13b-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-13b-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-34b-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-34b-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-70b-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-70b-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-7b-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-7b-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-13b-Python-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-13b-Python-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-34b-Python-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-34b-Python-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-70b-Python-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-70b-Python-hf","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/codellama/CodeLlama-7b-Python-hf":{"mode":"chat","base_model":"together/codellama/CodeLlama-7b-Python-hf","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NumbersStation/nsql-llama-2-7B":{"mode":"chat","base_model":"together/NumbersStation/nsql-llama-2-7B","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Phind/Phind-CodeLlama-34B-Python-v1":{"mode":"chat","base_model":"together/Phind/Phind-CodeLlama-34B-Python-v1","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Phind/Phind-CodeLlama-34B-v2":{"mode":"chat","base_model":"together/Phind/Phind-CodeLlama-34B-v2","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/WizardLM/WizardCoder-Python-34B-V1.0":{"mode":"chat","base_model":"together/WizardLM/WizardCoder-Python-34B-V1.0","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/WizardLM/WizardCoder-15B-V1.0":{"mode":"chat","base_model":"together/WizardLM/WizardCoder-15B-V1.0","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/prompthero/openjourney":{"mode":"image_generation","base_model":"together/prompthero/openjourney","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/runwayml/stable-diffusion-v1-5":{"mode":"image_generation","base_model":"together/runwayml/stable-diffusion-v1-5","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/SG161222/Realistic_Vision_V3.0_VAE":{"mode":"image_generation","base_model":"together/SG161222/Realistic_Vision_V3.0_VAE","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/stabilityai/stable-diffusion-2-1":{"mode":"image_generation","base_model":"together/stabilityai/stable-diffusion-2-1","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/stabilityai/stable-diffusion-xl-base-1.0":{"mode":"image_generation","base_model":"together/stabilityai/stable-diffusion-xl-base-1.0","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/wavymulder/Analog-Diffusion":{"mode":"image_generation","base_model":"together/wavymulder/Analog-Diffusion","provider":"litellm","model_parameters":[{"id":"steps","label":"Steps","helpText":"Sampling steps is the number of iterations that the model runs to go from random noise to a recognizable image based on the text.","type":"number","default":20,"range":{"min":1,"max":100}},{"id":"seed","label":"Seed","helpText":"Random seed used for generation.","type":"number","default":42,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"results","label":"Results","helpText":"Number of result images to generate.","type":"number","default":1,"range":{"min":1,"max":12}},{"id":"height","label":"Height","helpText":"Height of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"width","label":"Width","helpText":"Width of the image to generate in number of pixels.","type":"number","default":256,"range":{"min":256,"max":1024,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Meta-Llama/Llama-Guard-7b":{"mode":"moderation","base_model":"together/Meta-Llama/Llama-Guard-7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/databricks/dolly-v2-12b":{"mode":"chat","base_model":"together/databricks/dolly-v2-12b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/databricks/dolly-v2-3b":{"mode":"chat","base_model":"together/databricks/dolly-v2-3b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/databricks/dolly-v2-7b":{"mode":"chat","base_model":"together/databricks/dolly-v2-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/HuggingFaceH4/zephyr-7b-beta":{"mode":"chat","base_model":"together/HuggingFaceH4/zephyr-7b-beta","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/HuggingFaceH4/starchat-alpha":{"mode":"chat","base_model":"together/HuggingFaceH4/starchat-alpha","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5":{"mode":"chat","base_model":"together/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/OpenAssistant/stablelm-7b-sft-v7-epoch-3":{"mode":"chat","base_model":"together/OpenAssistant/stablelm-7b-sft-v7-epoch-3","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Koala-13B":{"mode":"chat","base_model":"together/togethercomputer/Koala-13B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/Koala-7B":{"mode":"chat","base_model":"together/togethercomputer/Koala-7B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/lmsys/vicuna-13b-v1.3":{"mode":"chat","base_model":"together/lmsys/vicuna-13b-v1.3","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/lmsys/vicuna-7b-v1.3":{"mode":"chat","base_model":"together/lmsys/vicuna-7b-v1.3","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/lmsys/fastchat-t5-3b-v1.0":{"mode":"chat","base_model":"together/lmsys/fastchat-t5-3b-v1.0","provider":"litellm","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/mpt-30b-chat":{"mode":"chat","base_model":"together/togethercomputer/mpt-30b-chat","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/mpt-7b-chat":{"mode":"chat","base_model":"together/togethercomputer/mpt-7b-chat","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/guanaco-13b":{"mode":"chat","base_model":"together/togethercomputer/guanaco-13b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/guanaco-33b":{"mode":"chat","base_model":"together/togethercomputer/guanaco-33b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/guanaco-65b":{"mode":"chat","base_model":"together/togethercomputer/guanaco-65b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/togethercomputer/guanaco-7b":{"mode":"chat","base_model":"together/togethercomputer/guanaco-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/defog/sqlcoder":{"mode":"chat","base_model":"together/defog/sqlcoder","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/gpt-j-6b":{"mode":"chat","base_model":"together/EleutherAI/gpt-j-6b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/gpt-neox-20b":{"mode":"chat","base_model":"together/EleutherAI/gpt-neox-20b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/pythia-12b-v0":{"mode":"chat","base_model":"together/EleutherAI/pythia-12b-v0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/pythia-1b-v0":{"mode":"chat","base_model":"together/EleutherAI/pythia-1b-v0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/pythia-2.8b-v0":{"mode":"chat","base_model":"together/EleutherAI/pythia-2.8b-v0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/EleutherAI/pythia-6.9b":{"mode":"chat","base_model":"together/EleutherAI/pythia-6.9b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/google/flan-t5-xl":{"mode":"chat","base_model":"together/google/flan-t5-xl","provider":"litellm","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/google/flan-t5-xxl":{"mode":"chat","base_model":"together/google/flan-t5-xxl","provider":"litellm","max_input_tokens":512,"max_output_tokens":512,"max_tokens":512,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/huggyllama/llama-13b":{"mode":"chat","base_model":"together/huggyllama/llama-13b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/huggyllama/llama-30b":{"mode":"chat","base_model":"together/huggyllama/llama-30b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/huggyllama/llama-7b":{"mode":"chat","base_model":"together/huggyllama/llama-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mosaicml/mpt-7b":{"mode":"chat","base_model":"together/mosaicml/mpt-7b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/mosaicml/mpt-7b-instruct":{"mode":"chat","base_model":"together/mosaicml/mpt-7b-instruct","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-13b":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-13b","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NumbersStation/nsql-6B":{"mode":"chat","base_model":"together/NumbersStation/nsql-6B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/stabilityai/stablelm-base-alpha-3b":{"mode":"chat","base_model":"together/stabilityai/stablelm-base-alpha-3b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/stabilityai/stablelm-base-alpha-7b":{"mode":"chat","base_model":"together/stabilityai/stablelm-base-alpha-7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/bigcode/starcoder":{"mode":"chat","base_model":"together/bigcode/starcoder","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Replit-Code-v1 (3B)":{"mode":"chat","base_model":"together/Replit-Code-v1 (3B)","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Salesforce/codegen2-16B":{"mode":"chat","base_model":"together/Salesforce/codegen2-16B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/Salesforce/codegen2-7B":{"mode":"chat","base_model":"together/Salesforce/codegen2-7B","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"anthropic/claude-instant-1.2":{"mode":"chat","base_model":"anthropic/claude-instant-1.2","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"anthropic/claude-2.1":{"mode":"chat","base_model":"anthropic/claude-2.1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"anthropic/claude-3-opus-20240229":{"mode":"chat","base_model":"anthropic/claude-3-opus-20240229","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-sonnet-20240229":{"mode":"image_generation","base_model":"anthropic/claude-3-sonnet-20240229","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-5-sonnet-20241022":{"mode":"image_generation","base_model":"anthropic/claude-3-5-sonnet-20241022","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true},"anthropic/claude-3-5-sonnet-latest":{"mode":"image_generation","base_model":"anthropic/claude-3-5-sonnet-latest","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_system_messages":true},"anthropic/claude-3-haiku-20240307":{"mode":"image_generation","base_model":"anthropic/claude-3-haiku-20240307","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-3-7-sonnet-latest":{"mode":"image_generation","base_model":"anthropic/claude-3-7-sonnet-latest","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-sonnet-4-20250514":{"mode":"image_generation","base_model":"anthropic/claude-sonnet-4-20250514","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-opus-4-20250514":{"mode":"image_generation","base_model":"anthropic/claude-opus-4-20250514","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-opus-4-1-20250805":{"mode":"image_generation","base_model":"anthropic/claude-opus-4-1-20250805","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"anthropic/claude-sonnet-4-5-20250929":{"mode":"image_generation","base_model":"anthropic/claude-sonnet-4-5-20250929","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_web_search":true},"anthropic/claude-opus-4-5":{"mode":"image_generation","base_model":"anthropic/claude-opus-4-5","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"effort","label":"Effort","helpText":"Set the effort level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true,"supports_output_config":true},"anthropic/claude-opus-4-5-20251101":{"mode":"image_generation","base_model":"anthropic/claude-opus-4-5-20251101","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"effort","label":"Effort","helpText":"Set the effort level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_output_config":true},"azure/gpt-4-turbo-vision-128k":{"mode":"image_generation","base_model":"azure/gpt-4-turbo-vision-128k","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"azure/gpt-4o-audio-preview":{"mode":"chat","base_model":"azure/gpt-4o-audio-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"azure/gpt-4o-audio-preview-2025-06-03":{"mode":"chat","base_model":"azure/gpt-4o-audio-preview-2025-06-03","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"azure/gpt-4o-audio-preview-2024-10-01":{"mode":"chat","base_model":"azure/gpt-4o-audio-preview-2024-10-01","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"azure/gpt-4o-mini-audio-preview":{"mode":"chat","base_model":"azure/gpt-4o-mini-audio-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]},{"id":"modalities","label":"Modalities","helpText":"The modalities to use for the model.","type":"select","multiple":true,"hidden":true,"default":["text"],"options":[{"label":"Text","value":"text"},{"label":"Audio","value":"audio"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_vision":true,"source":"merged_from_llm_models_csv"},"azure/phi-4":{"mode":"chat","base_model":"azure/phi-4","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192},{"id":"imageDetail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"presencePenalty","label":"Presence penalty","helpText":"How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"azure/gpt-5.1-chat-latest":{"mode":"image_generation","base_model":"azure/gpt-5.1-chat-latest","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"disabled":true,"disabledText":"Temperature value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"disabled":true,"disabledText":"Top P value is fixed at 1 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"disabled":true,"disabledText":"Frequency Penalty value is fixed at 0 and cannot be changed at the moment for this model.","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":128000,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_vision":true},"azure/DeepSeek-V3.1":{"mode":"chat","base_model":"azure/DeepSeek-V3.1","deprecation_date":"2026-07-13","source":"merged_from_llm_models_csv","provider":"litellm","is_deprecated":true,"max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true},"azure/DeepSeek-R1-0528":{"mode":"chat","base_model":"azure/DeepSeek-R1-0528","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_system_messages":true,"supports_tool_choice":false},"azure/DeepSeek-V3-0324":{"mode":"chat","base_model":"azure/DeepSeek-V3-0324","deprecation_date":"2026-07-13","source":"merged_from_llm_models_csv","provider":"litellm","is_deprecated":true,"max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true},"azure/DeepSeek-V3":{"mode":"chat","base_model":"azure/DeepSeek-V3","deprecation_date":"2025-08-31","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"provider":"litellm","is_deprecated":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true},"azure/DeepSeek-R1":{"mode":"chat","base_model":"azure/DeepSeek-R1","deprecation_date":"2026-08-13","source":"merged_from_llm_models_csv","provider":"litellm","is_deprecated":true,"max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"responseFormat","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}]},{"id":"frequencyPenalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true},"google/gemini-1.5-pro":{"mode":"image_generation","base_model":"google/gemini-1.5-pro","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-pro-latest":{"mode":"image_generation","base_model":"google/gemini-1.5-pro-latest","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-pro-002":{"mode":"image_generation","base_model":"google/gemini-1.5-pro-002","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash":{"mode":"image_generation","base_model":"google/gemini-1.5-flash","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash-latest":{"mode":"image_generation","base_model":"google/gemini-1.5-flash-latest","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash-002":{"mode":"image_generation","base_model":"google/gemini-1.5-flash-002","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash-8b":{"mode":"image_generation","base_model":"google/gemini-1.5-flash-8b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash-8b-latest":{"mode":"image_generation","base_model":"google/gemini-1.5-flash-8b-latest","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-1.5-flash-8b-001":{"mode":"image_generation","base_model":"google/gemini-1.5-flash-8b-001","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash-lite-preview-02-05":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-lite-preview-02-05","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-pro-preview-03-25":{"mode":"image_generation","base_model":"google/gemini-2.5-pro-preview-03-25","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash":{"mode":"chat","base_model":"google/gemini-2.0-flash","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-pro":{"mode":"chat","base_model":"google/gemini-2.5-pro","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_web_search":true},"google/gemini-2.5-flash-preview-05-20":{"mode":"chat","base_model":"google/gemini-2.5-flash-preview-05-20","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-flash-preview-04-17":{"mode":"chat","base_model":"google/gemini-2.5-flash-preview-04-17","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.5-flash":{"mode":"chat","base_model":"google/gemini-2.5-flash","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":1,"max":20,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"merged_from_llm_models_csv","supports_web_search":true,"supports_image_size":false},"google/gemini-2.0-flash-exp":{"mode":"chat","base_model":"google/gemini-2.0-flash-exp","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash-001":{"mode":"chat","base_model":"google/gemini-2.0-flash-001","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash-lite":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-lite","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-flash-lite-preview":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-lite-preview","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-pro-exp":{"mode":"image_generation","base_model":"google/gemini-2.0-pro-exp","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"google/gemini-2.0-pro-exp-02-05":{"mode":"image_generation","base_model":"google/gemini-2.0-pro-exp-02-05","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv"},"google/gemini-3-pro-preview":{"mode":"image_generation","base_model":"google/gemini-3-pro-preview","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"thinking_level","label":"Thinking Level","helpText":"Set the thinking level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"High","value":"high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true},"google/gemini-2.0-flash-thinking-exp-01-21":{"mode":"image_generation","base_model":"google/gemini-2.0-flash-thinking-exp-01-21","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":10,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01}},{"id":"topK","label":"Top K","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":1,"max":100,"step":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"groq/mixtral-8x7b":{"mode":"chat","base_model":"groq/mixtral-8x7b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"groq/llama3-8b-8192":{"mode":"chat","base_model":"groq/llama3-8b-8192","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"groq/llama3-70b-8192":{"mode":"chat","base_model":"groq/llama3-70b-8192","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"groq/llama-3.1-70b-versatile":{"mode":"chat","base_model":"groq/llama-3.1-70b-versatile","provider":"litellm","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"groq/llama-3.1-405b-reasoning":{"mode":"chat","base_model":"groq/llama-3.1-405b-reasoning","provider":"litellm","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"groq/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"groq/deepseek-r1-distill-llama-70b","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"together/Qwen/Qwen1.5-72B-Chat":{"mode":"chat","base_model":"together/Qwen/Qwen1.5-72B-Chat","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/NousResearch/Nous-Hermes-2-Mistral-7B-DPO":{"mode":"chat","base_model":"together/NousResearch/Nous-Hermes-2-Mistral-7B-DPO","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/deepseek-ai/deepseek-coder-33b-instruct":{"mode":"chat","base_model":"together/deepseek-ai/deepseek-coder-33b-instruct","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"together/deepseek-ai/DeepSeek-R1","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"together/google/gemma-7b-it":{"mode":"chat","base_model":"together/google/gemma-7b-it","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo":{"mode":"chat","base_model":"together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/meta-llama/Llama-3-8b-chat-hf":{"mode":"chat","base_model":"together/meta-llama/Llama-3-8b-chat-hf","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"image_generation","base_model":"together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_function_calling":true},"together/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"image_generation","base_model":"together/meta-llama/Llama-4-Scout-17B-16E-Instruct","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"min_p","label":"Min P","helpText":"A number between 0 and 1 that can be used as an alternative to top_p and top-k","type":"number","default":0,"range":{"min":0,"max":1}},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo":{"mode":"chat","base_model":"together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo":{"mode":"chat","base_model":"together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo":{"mode":"chat","base_model":"together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo":{"mode":"image_generation","base_model":"together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"together/meta-llama/Llama-3.2-3B-Instruct-Turbo":{"mode":"chat","base_model":"together/meta-llama/Llama-3.2-3B-Instruct-Turbo","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"cohere/command-nightly":{"mode":"chat","base_model":"cohere/command-nightly","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":128000,"step":10}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"cohere/command-light":{"mode":"chat","base_model":"cohere/command-light","provider":"litellm","max_input_tokens":4000,"max_output_tokens":4000,"max_tokens":4000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"cohere/command-light-nightly":{"mode":"chat","base_model":"cohere/command-light-nightly","provider":"litellm","max_input_tokens":4000,"max_output_tokens":4000,"max_tokens":4000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.3,"range":{"min":0,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_input_tokens","label":"Max input tokens","helpText":"The maximum number of input tokens to send to the model. If not specified, max_input_tokens is the model's context length limit minus a small buffer.","type":"number","default":1000,"range":{"min":1,"max":4000,"step":1}},{"id":"p","label":"P","helpText":"Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step. If both k and p are enabled, p acts after k.","type":"number","default":0.75,"range":{"min":0.01,"max":0.99,"step":0.01}},{"id":"k","label":"K","helpText":"Ensures only the top k most likely tokens are considered for generation at each step.","type":"number","default":10,"range":{"min":0,"max":500,"step":1}},{"id":"citation_quality","label":"Citation quality","helpText":"Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want 'accurate' results or 'fast' results.","type":"select","options":[{"label":"Accurate","value":"accurate"},{"label":"Fast","value":"fast"}]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Used to reduce repetitiveness of generated tokens. The higher the value, the stronger a penalty is applied to previously present tokens, proportional to how many times they have already appeared in the prompt or prior generation.","type":"number","range":{"min":0.1,"max":1,"step":0.1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"If specified, the backend will make a best effort to sample tokens deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"force_single_step","label":"Force single step","helpText":"Forces the chat to be single step.","type":"boolean","default":false},{"id":"preamble","label":"Preamble","helpText":"A short set of instructions given to an AI language model at the start of a conversation. It tells the AI how to behave and respond, like setting the rules of the game before playing. Users can change these instructions to make the AI act differently.","type":"text"},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"bedrock/anthropic.claude-3-haiku-20240307-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-3-haiku-20240307-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_cache_point":false,"supports_pdf_input":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_assistant_prefill":true},"bedrock/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-3-5-sonnet-20240620-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_pdf_input":true,"supports_response_schema":true,"supports_vision":true,"supports_assistant_prefill":true,"supports_cache_point":false},"bedrock/anthropic.claude-3-5-sonnet-20241022-v2:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-3-5-sonnet-20241022-v2:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_cache_point":false},"bedrock/anthropic.claude-v2":{"mode":"chat","base_model":"bedrock/anthropic.claude-v2","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"bedrock/anthropic.claude-instant-v1":{"mode":"chat","base_model":"bedrock/anthropic.claude-instant-v1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_assistant_prefill":true,"supports_cache_point":false},"bedrock/anthropic.claude-v2:1":{"mode":"chat","base_model":"bedrock/anthropic.claude-v2:1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_assistant_prefill":true,"supports_cache_point":false},"bedrock/anthropic.claude-3-sonnet-20240229-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-3-sonnet-20240229-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_cache_point":false,"supports_pdf_input":true,"supports_response_schema":true,"supports_vision":true,"supports_assistant_prefill":true},"bedrock/anthropic.claude-3-opus-20240229-v1:0":{"mode":"chat","base_model":"bedrock/anthropic.claude-3-opus-20240229-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true,"supports_vision":true,"supports_cache_point":false},"bedrock/anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-3-7-sonnet-20250219-v1:0","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"bedrock/anthropic.claude-opus-4-20250514-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-opus-4-20250514-v1:0","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true},"bedrock/anthropic.claude-sonnet-4-20250514-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-sonnet-4-20250514-v1:0","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false},"bedrock/anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-sonnet-4-5-20250929-v1:0","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_search":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false},"bedrock/anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"image_generation","base_model":"bedrock/anthropic.claude-haiku-4-5-20251001-v1:0","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_search":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_web_search":true},"bedrock/claude-opus-4-5":{"mode":"image_generation","base_model":"bedrock/claude-opus-4-5","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"effort","label":"Effort","helpText":"Set the effort level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_output_config":true,"supports_web_search":true},"bedrock/claude-opus-4-5-20251101":{"mode":"image_generation","base_model":"bedrock/claude-opus-4-5-20251101","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"disabledCondition":{"paramId":"thinking","operator":"eq","value":true,"setValue":1,"disabledText":"Temperature is disabled when extended thinking is enabled."},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"effort","label":"Effort","helpText":"Set the effort level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"thinking","label":"Extended thinking","helpText":"Enable the extended thinking parameter","type":"boolean","default":true},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_web_search":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_output_config":true},"bedrock/amazon.nova-lite-v1:0":{"mode":"image_generation","base_model":"bedrock/amazon.nova-lite-v1:0","provider":"litellm","max_input_tokens":5120,"max_output_tokens":5120,"max_tokens":5120,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":5120,"range":{"min":1,"max":5120},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true},"bedrock/amazon.nova-micro-v1:0":{"mode":"chat","base_model":"bedrock/amazon.nova-micro-v1:0","provider":"litellm","max_input_tokens":5120,"max_output_tokens":5120,"max_tokens":5120,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":5120,"range":{"min":1,"max":5120},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_prompt_caching":true,"supports_response_schema":true},"bedrock/amazon.nova-pro-v1:0":{"mode":"image_generation","base_model":"bedrock/amazon.nova-pro-v1:0","provider":"litellm","max_input_tokens":5120,"max_output_tokens":5120,"max_tokens":5120,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":5120,"range":{"min":1,"max":5120},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_service_tier":true},"bedrock/amazon.titan-text-lite-v1":{"mode":"chat","base_model":"bedrock/amazon.titan-text-lite-v1","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/amazon.titan-text-express-v1":{"mode":"chat","base_model":"bedrock/amazon.titan-text-express-v1","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/amazon.titan-text-premier-v1:0":{"mode":"chat","base_model":"bedrock/amazon.titan-text-premier-v1:0","provider":"litellm","max_input_tokens":3000,"max_output_tokens":3000,"max_tokens":3000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"stopSequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3000,"range":{"min":1,"max":3000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/meta.llama3-70b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-70b-instruct-v1:0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_pdf_input":true},"bedrock/meta.llama3-8b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-8b-instruct-v1:0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_pdf_input":true},"bedrock/meta.llama3-1-70b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-1-70b-instruct-v1:0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false},"bedrock/meta.llama3-1-8b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-1-8b-instruct-v1:0","provider":"litellm","max_input_tokens":2048,"max_output_tokens":2048,"max_tokens":2048,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false},"bedrock/meta.llama3-2-90b-instruct-v1:0":{"mode":"image_generation","base_model":"bedrock/meta.llama3-2-90b-instruct-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true},"bedrock/meta.llama3-2-11b-instruct-v1:0":{"mode":"image_generation","base_model":"bedrock/meta.llama3-2-11b-instruct-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false,"supports_vision":true},"bedrock/meta.llama3-2-3b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-2-3b-instruct-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false},"bedrock/meta.llama3-2-1b-instruct-v1:0":{"mode":"chat","base_model":"bedrock/meta.llama3-2-1b-instruct-v1:0","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_tool_choice":false},"bedrock/deepseek.r1-v1:0":{"mode":"chat","base_model":"bedrock/deepseek.r1-v1:0","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_tool_choice":false},"bedrock/deepseek-llm-r1-distill-qwen-7b":{"mode":"chat","base_model":"bedrock/deepseek-llm-r1-distill-qwen-7b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"bedrock/deepseek-llm-r1-distill-qwen-32b":{"mode":"chat","base_model":"bedrock/deepseek-llm-r1-distill-qwen-32b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"bedrock/deepseek-llm-r1-distill-qwen-14b":{"mode":"chat","base_model":"bedrock/deepseek-llm-r1-distill-qwen-14b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"bedrock/deepseek-llm-r1-distill-llama-8b":{"mode":"chat","base_model":"bedrock/deepseek-llm-r1-distill-llama-8b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"bedrock/deepseek-llm-r1-distill-llama-70b":{"mode":"chat","base_model":"bedrock/deepseek-llm-r1-distill-llama-70b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-coder-7b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-coder-7b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-coder-32b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-coder-32b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-7b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-7b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-72b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-72b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-32b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-32b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/huggingface-llm-qwen2-5-14b-instruct":{"mode":"chat","base_model":"bedrock/huggingface-llm-qwen2-5-14b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"bedrock/openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"bedrock/openai.gpt-oss-120b-1:0","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":false,"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true},"bedrock/openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"bedrock/openai.gpt-oss-20b-1:0","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":false,"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true},"mistral/ministral-8b-latest":{"mode":"chat","base_model":"mistral/ministral-8b-latest","max_input_tokens":131000,"max_output_tokens":131000,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"litellm","max_tokens":131000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131000,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/pixtral-12b":{"mode":"image_generation","base_model":"mistral/pixtral-12b","provider":"litellm","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"image_detail","label":"Image Detail","helpText":"By controlling the detail parameter, you have control over how the model processes the image and generates its textual understanding. You can override this value at attachment level.","type":"select","accesorKey":"detail","default":{"detail":"auto"},"options":[{"label":"Auto","value":"auto"},{"label":"Low","value":"low"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"merged_from_llm_models_csv","supports_vision":true},"mistral/mistral-saba-latest":{"mode":"chat","base_model":"mistral/mistral-saba-latest","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32000,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"mistral/ministral-3b-latest":{"mode":"chat","base_model":"mistral/ministral-3b-latest","max_input_tokens":131000,"max_output_tokens":131000,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"litellm","max_tokens":131000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"maxTokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536},{"id":"topP","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0},"default":0.9},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131000,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/yi-01-ai/models/yi-large":{"mode":"chat","base_model":"fireworks/accounts/yi-01-ai/models/yi-large","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_response_schema":true,"supports_tool_choice":false},"fireworks_ai/accounts/sentientfoundation/models/dobby-unhinged-llama-3-3-70b-new":{"mode":"chat","base_model":"fireworks/accounts/sentientfoundation/models/dobby-unhinged-llama-3-3-70b-new","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/alpha":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/alpha","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"fireworks_ai/accounts/fireworks/models/moa":{"mode":"chat","base_model":"fireworks/accounts/fireworks/models/moa","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/openrouter/horizon-beta":{"mode":"image_generation","base_model":"openrouter/openrouter/horizon-beta","provider":"litellm","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/codestral-2508":{"mode":"chat","base_model":"openrouter/mistralai/codestral-2508","provider":"litellm","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-30b-a3b-instruct-2507":{"mode":"chat","base_model":"openrouter/qwen/qwen3-30b-a3b-instruct-2507","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/z-ai/glm-4.5":{"mode":"chat","base_model":"openrouter/z-ai/glm-4.5","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/z-ai/glm-4.5-air:free":{"mode":"chat","base_model":"openrouter/z-ai/glm-4.5-air:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/z-ai/glm-4.5-air":{"mode":"chat","base_model":"openrouter/z-ai/glm-4.5-air","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/z-ai/glm-4-32b":{"mode":"chat","base_model":"openrouter/z-ai/glm-4-32b","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.5-flash-lite":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.5-flash-lite","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/moonshotai/kimi-k2:free":{"mode":"chat","base_model":"openrouter/moonshotai/kimi-k2:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/moonshotai/kimi-k2":{"mode":"chat","base_model":"openrouter/moonshotai/kimi-k2","provider":"litellm","max_input_tokens":63000,"max_output_tokens":63000,"max_tokens":63000,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":63000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/thudm/glm-4.1v-9b-thinking":{"mode":"image_generation","base_model":"openrouter/thudm/glm-4.1v-9b-thinking","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/devstral-medium":{"mode":"chat","base_model":"openrouter/mistralai/devstral-medium","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/devstral-small":{"mode":"chat","base_model":"openrouter/mistralai/devstral-small","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_response_schema":true},"openrouter/cognitivecomputations/dolphin-mistral-24b-venice-edition:free":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin-mistral-24b-venice-edition:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3n-e2b-it:free":{"mode":"chat","base_model":"openrouter/google/gemma-3n-e2b-it:free","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/tencent/hunyuan-a13b-instruct:free":{"mode":"chat","base_model":"openrouter/tencent/hunyuan-a13b-instruct:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/tencent/hunyuan-a13b-instruct":{"mode":"chat","base_model":"openrouter/tencent/hunyuan-a13b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/tngtech/deepseek-r1t2-chimera:free":{"mode":"chat","base_model":"openrouter/tngtech/deepseek-r1t2-chimera:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/morph/morph-v3-large":{"mode":"chat","base_model":"openrouter/morph/morph-v3-large","provider":"litellm","max_input_tokens":81920,"max_output_tokens":81920,"max_tokens":81920,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":81920},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/morph/morph-v3-fast":{"mode":"chat","base_model":"openrouter/morph/morph-v3-fast","provider":"litellm","max_input_tokens":81920,"max_output_tokens":81920,"max_tokens":81920,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":81920},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/baidu/ernie-4.5-300b-a47b":{"mode":"chat","base_model":"openrouter/baidu/ernie-4.5-300b-a47b","provider":"litellm","max_input_tokens":123000,"max_output_tokens":123000,"max_tokens":123000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":123000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/anubis-70b-v1.1":{"mode":"chat","base_model":"openrouter/thedrummer/anubis-70b-v1.1","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/inception/mercury":{"mode":"chat","base_model":"openrouter/inception/mercury","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-small-3.2-24b-instruct:free":{"mode":"image_generation","base_model":"openrouter/mistralai/mistral-small-3.2-24b-instruct:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/minimax/minimax-m1":{"mode":"chat","base_model":"openrouter/minimax/minimax-m1","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.5-flash-lite-preview-06-17":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.5-flash-lite-preview-06-17","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/moonshotai/kimi-dev-72b:free":{"mode":"chat","base_model":"openrouter/moonshotai/kimi-dev-72b:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/o3-pro":{"mode":"image_generation","base_model":"openrouter/openai/o3-pro","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-3-mini":{"mode":"chat","base_model":"openrouter/x-ai/grok-3-mini","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_pdf_input":false},"openrouter/x-ai/grok-3":{"mode":"chat","base_model":"openrouter/x-ai/grok-3","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_pdf_input":false},"openrouter/mistralai/magistral-small-2506":{"mode":"chat","base_model":"openrouter/mistralai/magistral-small-2506","provider":"litellm","max_input_tokens":40000,"max_output_tokens":40000,"max_tokens":40000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/magistral-medium-2506":{"mode":"chat","base_model":"openrouter/mistralai/magistral-medium-2506","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/magistral-medium-2506:thinking":{"mode":"chat","base_model":"openrouter/mistralai/magistral-medium-2506:thinking","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.5-pro-preview":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.5-pro-preview","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-qwen-7b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-qwen-7b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-0528-qwen3-8b:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-0528-qwen3-8b:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-0528-qwen3-8b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-0528-qwen3-8b","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true},"openrouter/deepseek/deepseek-r1-0528:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-0528:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/sarvamai/sarvam-m:free":{"mode":"chat","base_model":"openrouter/sarvamai/sarvam-m:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/valkyrie-49b-v1":{"mode":"chat","base_model":"openrouter/thedrummer/valkyrie-49b-v1","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/devstral-small-2505:free":{"mode":"chat","base_model":"openrouter/mistralai/devstral-small-2505:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/devstral-small-2505":{"mode":"chat","base_model":"openrouter/mistralai/devstral-small-2505","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3n-e4b-it:free":{"mode":"chat","base_model":"openrouter/google/gemma-3n-e4b-it:free","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3n-e4b-it":{"mode":"chat","base_model":"openrouter/google/gemma-3n-e4b-it","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/codex-mini":{"mode":"image_generation","base_model":"openrouter/openai/codex-mini","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true},"openrouter/nousresearch/deephermes-3-mistral-24b-preview":{"mode":"chat","base_model":"openrouter/nousresearch/deephermes-3-mistral-24b-preview","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-medium-3":{"mode":"image_generation","base_model":"openrouter/mistralai/mistral-medium-3","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.5-pro-preview-05-06":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.5-pro-preview-05-06","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/arcee-ai/spotlight":{"mode":"image_generation","base_model":"openrouter/arcee-ai/spotlight","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/arcee-ai/maestro-reasoning":{"mode":"chat","base_model":"openrouter/arcee-ai/maestro-reasoning","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/arcee-ai/virtuoso-large":{"mode":"chat","base_model":"openrouter/arcee-ai/virtuoso-large","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/arcee-ai/coder-large":{"mode":"chat","base_model":"openrouter/arcee-ai/coder-large","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-4-reasoning-plus":{"mode":"chat","base_model":"openrouter/microsoft/phi-4-reasoning-plus","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":false},"openrouter/inception/mercury-coder":{"mode":"chat","base_model":"openrouter/inception/mercury-coder","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-4b:free":{"mode":"chat","base_model":"openrouter/qwen/qwen3-4b:free","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/opengvlab/internvl3-14b":{"mode":"image_generation","base_model":"openrouter/opengvlab/internvl3-14b","provider":"litellm","max_input_tokens":12288,"max_output_tokens":12288,"max_tokens":12288,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":12288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-prover-v2":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-prover-v2","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-guard-4-12b":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-guard-4-12b","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-30b-a3b:free":{"mode":"chat","base_model":"openrouter/qwen/qwen3-30b-a3b:free","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-30b-a3b":{"mode":"chat","base_model":"openrouter/qwen/qwen3-30b-a3b","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-8b:free":{"mode":"chat","base_model":"openrouter/qwen/qwen3-8b:free","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-8b":{"mode":"chat","base_model":"openrouter/qwen/qwen3-8b","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_response_schema":true},"openrouter/qwen/qwen3-14b:free":{"mode":"chat","base_model":"openrouter/qwen/qwen3-14b:free","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-14b":{"mode":"chat","base_model":"openrouter/qwen/qwen3-14b","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-32b":{"mode":"chat","base_model":"openrouter/qwen/qwen3-32b","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-235b-a22b:free":{"mode":"chat","base_model":"openrouter/qwen/qwen3-235b-a22b:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen3-235b-a22b":{"mode":"chat","base_model":"openrouter/qwen/qwen3-235b-a22b","provider":"litellm","max_input_tokens":40960,"max_output_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/tngtech/deepseek-r1t-chimera:free":{"mode":"chat","base_model":"openrouter/tngtech/deepseek-r1t-chimera:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/tngtech/deepseek-r1t-chimera":{"mode":"chat","base_model":"openrouter/tngtech/deepseek-r1t-chimera","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/microsoft/mai-ds-r1:free":{"mode":"chat","base_model":"openrouter/microsoft/mai-ds-r1:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/microsoft/mai-ds-r1":{"mode":"chat","base_model":"openrouter/microsoft/mai-ds-r1","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/thudm/glm-z1-32b:free":{"mode":"chat","base_model":"openrouter/thudm/glm-z1-32b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/thudm/glm-4-32b":{"mode":"chat","base_model":"openrouter/thudm/glm-4-32b","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/o4-mini-high":{"mode":"image_generation","base_model":"openrouter/openai/o4-mini-high","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/openai/o3":{"mode":"image_generation","base_model":"openrouter/openai/o3","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/o4-mini":{"mode":"image_generation","base_model":"openrouter/openai/o4-mini","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"openrouter/shisa-ai/shisa-v2-llama3.3-70b:free":{"mode":"chat","base_model":"openrouter/shisa-ai/shisa-v2-llama3.3-70b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/shisa-ai/shisa-v2-llama3.3-70b":{"mode":"chat","base_model":"openrouter/shisa-ai/shisa-v2-llama3.3-70b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/eleutherai/llemma_7b":{"mode":"chat","base_model":"openrouter/eleutherai/llemma_7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/alfredpros/codellama-7b-instruct-solidity":{"mode":"chat","base_model":"openrouter/alfredpros/codellama-7b-instruct-solidity","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/arliai/qwq-32b-arliai-rpr-v1:free":{"mode":"chat","base_model":"openrouter/arliai/qwq-32b-arliai-rpr-v1:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/arliai/qwq-32b-arliai-rpr-v1":{"mode":"chat","base_model":"openrouter/arliai/qwq-32b-arliai-rpr-v1","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/agentica-org/deepcoder-14b-preview:free":{"mode":"chat","base_model":"openrouter/agentica-org/deepcoder-14b-preview:free","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/agentica-org/deepcoder-14b-preview":{"mode":"chat","base_model":"openrouter/agentica-org/deepcoder-14b-preview","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/moonshotai/kimi-vl-a3b-thinking:free":{"mode":"image_generation","base_model":"openrouter/moonshotai/kimi-vl-a3b-thinking:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/moonshotai/kimi-vl-a3b-thinking":{"mode":"image_generation","base_model":"openrouter/moonshotai/kimi-vl-a3b-thinking","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-3-mini-beta":{"mode":"chat","base_model":"openrouter/x-ai/grok-3-mini-beta","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-3-beta":{"mode":"chat","base_model":"openrouter/x-ai/grok-3-beta","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/nvidia/llama-3.3-nemotron-super-49b-v1":{"mode":"chat","base_model":"openrouter/nvidia/llama-3.3-nemotron-super-49b-v1","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/nvidia/llama-3.1-nemotron-ultra-253b-v1:free":{"mode":"chat","base_model":"openrouter/nvidia/llama-3.1-nemotron-ultra-253b-v1:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/nvidia/llama-3.1-nemotron-ultra-253b-v1":{"mode":"chat","base_model":"openrouter/nvidia/llama-3.1-nemotron-ultra-253b-v1","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-4-maverick":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-4-maverick","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-4-scout":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-4-scout","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-v3-base":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-v3-base","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/scb10x/llama3.1-typhoon2-70b-instruct":{"mode":"chat","base_model":"openrouter/scb10x/llama3.1-typhoon2-70b-instruct","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.5-pro-exp-03-25":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.5-pro-exp-03-25","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen2.5-vl-32b-instruct:free":{"mode":"image_generation","base_model":"openrouter/qwen/qwen2.5-vl-32b-instruct:free","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen2.5-vl-32b-instruct":{"mode":"image_generation","base_model":"openrouter/qwen/qwen2.5-vl-32b-instruct","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-chat-v3-0324:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-chat-v3-0324:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/featherless/qwerky-72b:free":{"mode":"chat","base_model":"openrouter/featherless/qwerky-72b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/o1-pro":{"mode":"image_generation","base_model":"openrouter/openai/o1-pro","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_native_streaming":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true},"openrouter/mistralai/mistral-small-3.1-24b-instruct:free":{"mode":"image_generation","base_model":"openrouter/mistralai/mistral-small-3.1-24b-instruct:free","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-4b-it:free":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-4b-it:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-4b-it":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-4b-it","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/ai21/jamba-1.6-large":{"mode":"chat","base_model":"openrouter/ai21/jamba-1.6-large","provider":"litellm","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/ai21/jamba-1.6-mini":{"mode":"chat","base_model":"openrouter/ai21/jamba-1.6-mini","provider":"litellm","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-12b-it:free":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-12b-it:free","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-12b-it":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-12b-it","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cohere/command-a":{"mode":"chat","base_model":"openrouter/cohere/command-a","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o-mini-search-preview":{"mode":"chat","base_model":"openrouter/openai/gpt-4o-mini-search-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"openrouter/openai/gpt-4o-search-preview":{"mode":"chat","base_model":"openrouter/openai/gpt-4o-search-preview","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"openrouter/rekaai/reka-flash-3:free":{"mode":"chat","base_model":"openrouter/rekaai/reka-flash-3:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-27b-it:free":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-27b-it:free","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-3-27b-it":{"mode":"image_generation","base_model":"openrouter/google/gemma-3-27b-it","provider":"litellm","max_input_tokens":96000,"max_output_tokens":96000,"max_tokens":96000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":96000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/anubis-pro-105b-v1":{"mode":"chat","base_model":"openrouter/thedrummer/anubis-pro-105b-v1","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/skyfall-36b-v2":{"mode":"chat","base_model":"openrouter/thedrummer/skyfall-36b-v2","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-4-multimodal-instruct":{"mode":"image_generation","base_model":"openrouter/microsoft/phi-4-multimodal-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/perplexity/sonar-reasoning-pro":{"mode":"image_generation","base_model":"openrouter/perplexity/sonar-reasoning-pro","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_reasoning":true,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/perplexity/sonar-pro":{"mode":"image_generation","base_model":"openrouter/perplexity/sonar-pro","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/perplexity/sonar-deep-research":{"mode":"chat","base_model":"openrouter/perplexity/sonar-deep-research","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_reasoning":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwq-32b:free":{"mode":"chat","base_model":"openrouter/qwen/qwq-32b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwq-32b":{"mode":"chat","base_model":"openrouter/qwen/qwq-32b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/nousresearch/deephermes-3-llama-3-8b-preview:free":{"mode":"chat","base_model":"openrouter/nousresearch/deephermes-3-llama-3-8b-preview:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.0-flash-lite-001":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.0-flash-lite-001","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.7-sonnet:thinking":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.7-sonnet:thinking","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.7-sonnet:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.7-sonnet:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/perplexity/r1-1776":{"mode":"chat","base_model":"openrouter/perplexity/r1-1776","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-saba":{"mode":"chat","base_model":"openrouter/mistralai/mistral-saba","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cognitivecomputations/dolphin3.0-r1-mistral-24b:free":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin3.0-r1-mistral-24b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/cognitivecomputations/dolphin3.0-r1-mistral-24b":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin3.0-r1-mistral-24b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/cognitivecomputations/dolphin3.0-mistral-24b:free":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin3.0-mistral-24b:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cognitivecomputations/dolphin3.0-mistral-24b":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin3.0-mistral-24b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-guard-3-8b":{"mode":"chat","base_model":"openrouter/meta-llama/llama-guard-3-8b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false},"openrouter/deepseek/deepseek-r1-distill-llama-8b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-llama-8b","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/aion-labs/aion-1.0":{"mode":"chat","base_model":"openrouter/aion-labs/aion-1.0","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/aion-labs/aion-1.0-mini":{"mode":"chat","base_model":"openrouter/aion-labs/aion-1.0-mini","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/aion-labs/aion-rp-llama-3.1-8b":{"mode":"chat","base_model":"openrouter/aion-labs/aion-rp-llama-3.1-8b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-vl-max":{"mode":"image_generation","base_model":"openrouter/qwen/qwen-vl-max","provider":"litellm","max_input_tokens":7500,"max_output_tokens":7500,"max_tokens":7500,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":7500},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-turbo":{"mode":"chat","base_model":"openrouter/qwen/qwen-turbo","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_reasoning":true},"openrouter/qwen/qwen2.5-vl-72b-instruct:free":{"mode":"image_generation","base_model":"openrouter/qwen/qwen2.5-vl-72b-instruct:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen2.5-vl-72b-instruct":{"mode":"image_generation","base_model":"openrouter/qwen/qwen2.5-vl-72b-instruct","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-plus":{"mode":"chat","base_model":"openrouter/qwen/qwen-plus","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-max":{"mode":"chat","base_model":"openrouter/qwen/qwen-max","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-qwen-1.5b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-qwen-1.5b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-small-24b-instruct-2501:free":{"mode":"chat","base_model":"openrouter/mistralai/mistral-small-24b-instruct-2501:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-small-24b-instruct-2501":{"mode":"chat","base_model":"openrouter/mistralai/mistral-small-24b-instruct-2501","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-qwen-32b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"openrouter/deepseek/deepseek-r1-distill-qwen-14b:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-qwen-14b:free","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-qwen-14b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-qwen-14b","provider":"litellm","max_input_tokens":64000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"openrouter/perplexity/sonar-reasoning":{"mode":"chat","base_model":"openrouter/perplexity/sonar-reasoning","provider":"litellm","max_input_tokens":127000,"max_output_tokens":127000,"max_tokens":127000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":127000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_reasoning":true,"supports_web_search":true},"openrouter/perplexity/sonar":{"mode":"image_generation","base_model":"openrouter/perplexity/sonar","provider":"litellm","max_input_tokens":127072,"max_output_tokens":127072,"max_tokens":127072,"supports_vision":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":127072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/liquid/lfm-7b":{"mode":"chat","base_model":"openrouter/liquid/lfm-7b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"openrouter/liquid/lfm-3b":{"mode":"chat","base_model":"openrouter/liquid/lfm-3b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-llama-70b:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-llama-70b:free","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/deepseek/deepseek-r1-distill-llama-70b":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1-distill-llama-70b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"openrouter/deepseek/deepseek-r1:free":{"mode":"chat","base_model":"openrouter/deepseek/deepseek-r1:free","provider":"litellm","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/minimax/minimax-01":{"mode":"image_generation","base_model":"openrouter/minimax/minimax-01","provider":"litellm","max_input_tokens":1000192,"max_output_tokens":1000192,"max_tokens":1000192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/codestral-2501":{"mode":"chat","base_model":"openrouter/mistralai/codestral-2501","provider":"litellm","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-4":{"mode":"chat","base_model":"openrouter/microsoft/phi-4","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/sao10k/l3.3-euryale-70b":{"mode":"chat","base_model":"openrouter/sao10k/l3.3-euryale-70b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-2-vision-1212":{"mode":"image_generation","base_model":"openrouter/x-ai/grok-2-vision-1212","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-2-1212":{"mode":"chat","base_model":"openrouter/x-ai/grok-2-1212","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/cohere/command-r7b-12-2024":{"mode":"chat","base_model":"openrouter/cohere/command-r7b-12-2024","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-2.0-flash-exp:free":{"mode":"image_generation","base_model":"openrouter/google/gemini-2.0-flash-exp:free","provider":"litellm","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.3-70b-instruct:free":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.3-70b-instruct:free","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.3-70b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.3-70b-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/amazon/nova-lite-v1":{"mode":"image_generation","base_model":"openrouter/amazon/nova-lite-v1","provider":"litellm","max_input_tokens":300000,"max_output_tokens":300000,"max_tokens":300000,"supports_function_calling":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":300000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/amazon/nova-micro-v1":{"mode":"chat","base_model":"openrouter/amazon/nova-micro-v1","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/amazon/nova-pro-v1":{"mode":"image_generation","base_model":"openrouter/amazon/nova-pro-v1","provider":"litellm","max_input_tokens":300000,"max_output_tokens":300000,"max_tokens":300000,"supports_function_calling":true,"supports_vision":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":300000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwq-32b-preview":{"mode":"chat","base_model":"openrouter/qwen/qwq-32b-preview","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o-2024-11-20":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4o-2024-11-20","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-large-2411":{"mode":"chat","base_model":"openrouter/mistralai/mistral-large-2411","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-large-2407":{"mode":"chat","base_model":"openrouter/mistralai/mistral-large-2407","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/pixtral-large-2411":{"mode":"image_generation","base_model":"openrouter/mistralai/pixtral-large-2411","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/x-ai/grok-vision-beta":{"mode":"image_generation","base_model":"openrouter/x-ai/grok-vision-beta","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/infermatic/mn-inferor-12b":{"mode":"chat","base_model":"openrouter/infermatic/mn-inferor-12b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-2.5-coder-32b-instruct:free":{"mode":"chat","base_model":"openrouter/qwen/qwen-2.5-coder-32b-instruct:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/raifle/sorcererlm-8x22b":{"mode":"chat","base_model":"openrouter/raifle/sorcererlm-8x22b","provider":"litellm","max_input_tokens":16000,"max_output_tokens":16000,"max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/unslopnemo-12b":{"mode":"chat","base_model":"openrouter/thedrummer/unslopnemo-12b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.5-haiku:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-haiku:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.5-haiku":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-haiku","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"openrouter/anthropic/claude-3.5-haiku-20241022":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-haiku-20241022","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthracite-org/magnum-v4-72b":{"mode":"chat","base_model":"openrouter/anthracite-org/magnum-v4-72b","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.5-sonnet:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-sonnet:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/ministral-8b":{"mode":"chat","base_model":"openrouter/mistralai/ministral-8b","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"openrouter/mistralai/ministral-3b":{"mode":"chat","base_model":"openrouter/mistralai/ministral-3b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true},"openrouter/qwen/qwen-2.5-7b-instruct":{"mode":"chat","base_model":"openrouter/qwen/qwen-2.5-7b-instruct","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/nvidia/llama-3.1-nemotron-70b-instruct":{"mode":"chat","base_model":"openrouter/nvidia/llama-3.1-nemotron-70b-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/inflection/inflection-3-productivity":{"mode":"chat","base_model":"openrouter/inflection/inflection-3-productivity","provider":"litellm","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/inflection/inflection-3-pi":{"mode":"chat","base_model":"openrouter/inflection/inflection-3-pi","provider":"litellm","max_input_tokens":8000,"max_output_tokens":8000,"max_tokens":8000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-flash-1.5-8b":{"mode":"image_generation","base_model":"openrouter/google/gemini-flash-1.5-8b","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/thedrummer/rocinante-12b":{"mode":"chat","base_model":"openrouter/thedrummer/rocinante-12b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/liquid/lfm-40b":{"mode":"chat","base_model":"openrouter/liquid/lfm-40b","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"openrouter/anthracite-org/magnum-v2-72b":{"mode":"chat","base_model":"openrouter/anthracite-org/magnum-v2-72b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-3b-instruct:free":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.2-3b-instruct:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-3b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.2-3b-instruct","provider":"litellm","max_input_tokens":20000,"max_output_tokens":20000,"max_tokens":20000,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":20000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-90b-vision-instruct":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-3.2-90b-vision-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-1b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.2-1b-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-11b-vision-instruct:free":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-3.2-11b-vision-instruct:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.2-11b-vision-instruct":{"mode":"image_generation","base_model":"openrouter/meta-llama/llama-3.2-11b-vision-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":false,"supports_reasoning":false},"openrouter/qwen/qwen-2.5-72b-instruct:free":{"mode":"chat","base_model":"openrouter/qwen/qwen-2.5-72b-instruct:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-2.5-72b-instruct":{"mode":"chat","base_model":"openrouter/qwen/qwen-2.5-72b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/neversleep/llama-3.1-lumimaid-8b":{"mode":"chat","base_model":"openrouter/neversleep/llama-3.1-lumimaid-8b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/o1-mini-2024-09-12":{"mode":"chat","base_model":"openrouter/openai/o1-mini-2024-09-12","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_vision":false},"openrouter/openai/o1-mini":{"mode":"chat","base_model":"openrouter/openai/o1-mini","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv","supports_system_messages":true},"openrouter/mistralai/pixtral-12b":{"mode":"image_generation","base_model":"openrouter/mistralai/pixtral-12b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"openrouter/cohere/command-r-08-2024":{"mode":"chat","base_model":"openrouter/cohere/command-r-08-2024","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cohere/command-r-plus-08-2024":{"mode":"chat","base_model":"openrouter/cohere/command-r-plus-08-2024","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-2.5-vl-7b-instruct":{"mode":"image_generation","base_model":"openrouter/qwen/qwen-2.5-vl-7b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/sao10k/l3.1-euryale-70b":{"mode":"chat","base_model":"openrouter/sao10k/l3.1-euryale-70b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-3.5-mini-128k-instruct":{"mode":"chat","base_model":"openrouter/microsoft/phi-3.5-mini-128k-instruct","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/nousresearch/hermes-3-llama-3.1-70b":{"mode":"chat","base_model":"openrouter/nousresearch/hermes-3-llama-3.1-70b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/nousresearch/hermes-3-llama-3.1-405b":{"mode":"chat","base_model":"openrouter/nousresearch/hermes-3-llama-3.1-405b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/chatgpt-4o-latest":{"mode":"image_generation","base_model":"openrouter/openai/chatgpt-4o-latest","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/sao10k/l3-lunaris-8b":{"mode":"chat","base_model":"openrouter/sao10k/l3-lunaris-8b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o-2024-08-06":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4o-2024-08-06","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.1-405b":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.1-405b","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.1-70b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.1-70b-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.1-405b-instruct:free":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.1-405b-instruct:free","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.1-405b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.1-405b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3.1-8b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3.1-8b-instruct","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-nemo:free":{"mode":"chat","base_model":"openrouter/mistralai/mistral-nemo:free","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-nemo":{"mode":"chat","base_model":"openrouter/mistralai/mistral-nemo","provider":"litellm","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o-mini-2024-07-18":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4o-mini-2024-07-18","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o-mini":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4o-mini","provider":"litellm","max_input_tokens":16384,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-2-27b-it":{"mode":"chat","base_model":"openrouter/google/gemma-2-27b-it","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"supports_response_schema":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-2-9b-it:free":{"mode":"chat","base_model":"openrouter/google/gemma-2-9b-it:free","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemma-2-9b-it":{"mode":"chat","base_model":"openrouter/google/gemma-2-9b-it","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.5-sonnet-20240620:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-sonnet-20240620:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3.5-sonnet-20240620":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3.5-sonnet-20240620","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/sao10k/l3-euryale-70b":{"mode":"chat","base_model":"openrouter/sao10k/l3-euryale-70b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cognitivecomputations/dolphin-mixtral-8x22b":{"mode":"chat","base_model":"openrouter/cognitivecomputations/dolphin-mixtral-8x22b","provider":"litellm","max_input_tokens":16000,"max_output_tokens":16000,"max_tokens":16000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/qwen/qwen-2-72b-instruct":{"mode":"chat","base_model":"openrouter/qwen/qwen-2-72b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-7b-instruct-v0.3":{"mode":"chat","base_model":"openrouter/mistralai/mistral-7b-instruct-v0.3","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/nousresearch/hermes-2-pro-llama-3-8b":{"mode":"chat","base_model":"openrouter/nousresearch/hermes-2-pro-llama-3-8b","provider":"litellm","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true,"supports_system_messages":true,"supports_response_schema":true},"openrouter/mistralai/mistral-7b-instruct:free":{"mode":"chat","base_model":"openrouter/mistralai/mistral-7b-instruct:free","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-3-mini-128k-instruct":{"mode":"chat","base_model":"openrouter/microsoft/phi-3-mini-128k-instruct","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/microsoft/phi-3-medium-128k-instruct":{"mode":"chat","base_model":"openrouter/microsoft/phi-3-medium-128k-instruct","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"openrouter/neversleep/llama-3-lumimaid-70b":{"mode":"chat","base_model":"openrouter/neversleep/llama-3-lumimaid-70b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/google/gemini-flash-1.5":{"mode":"image_generation","base_model":"openrouter/google/gemini-flash-1.5","provider":"litellm","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":131072,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-guard-2-8b":{"mode":"chat","base_model":"openrouter/meta-llama/llama-guard-2-8b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4o:extended":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4o:extended","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/sao10k/fimbulvetr-11b-v2":{"mode":"chat","base_model":"openrouter/sao10k/fimbulvetr-11b-v2","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/meta-llama/llama-3-8b-instruct":{"mode":"chat","base_model":"openrouter/meta-llama/llama-3-8b-instruct","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"openrouter/microsoft/wizardlm-2-8x22b":{"mode":"chat","base_model":"openrouter/microsoft/wizardlm-2-8x22b","provider":"litellm","max_input_tokens":65536,"max_output_tokens":65536,"max_tokens":65536,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":32768,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_system_messages":true,"supports_response_schema":true},"openrouter/google/gemini-pro-1.5":{"mode":"image_generation","base_model":"openrouter/google/gemini-pro-1.5","provider":"litellm","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":256,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4-turbo":{"mode":"image_generation","base_model":"openrouter/openai/gpt-4-turbo","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true},"openrouter/cohere/command-r-plus":{"mode":"chat","base_model":"openrouter/cohere/command-r-plus","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/cohere/command-r-plus-04-2024":{"mode":"chat","base_model":"openrouter/cohere/command-r-plus-04-2024","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/sophosympatheia/midnight-rose-70b":{"mode":"chat","base_model":"openrouter/sophosympatheia/midnight-rose-70b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cohere/command":{"mode":"chat","base_model":"openrouter/cohere/command","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/cohere/command-r":{"mode":"chat","base_model":"openrouter/cohere/command-r","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3-haiku:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3-haiku:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3-opus:beta":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3-opus:beta","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/anthropic/claude-3-opus":{"mode":"image_generation","base_model":"openrouter/anthropic/claude-3-opus","provider":"litellm","max_input_tokens":200000,"max_output_tokens":200000,"max_tokens":200000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":true},"openrouter/cohere/command-r-03-2024":{"mode":"chat","base_model":"openrouter/cohere/command-r-03-2024","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-3.5-turbo-0613":{"mode":"chat","base_model":"openrouter/openai/gpt-3.5-turbo-0613","provider":"litellm","max_input_tokens":4095,"max_output_tokens":4095,"max_tokens":4095,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4095},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4-turbo-preview":{"mode":"chat","base_model":"openrouter/openai/gpt-4-turbo-preview","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/nousresearch/nous-hermes-2-mixtral-8x7b-dpo":{"mode":"chat","base_model":"openrouter/nousresearch/nous-hermes-2-mixtral-8x7b-dpo","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/mistralai/mistral-small":{"mode":"chat","base_model":"openrouter/mistralai/mistral-small","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_reasoning":true,"supports_response_schema":true},"openrouter/mistralai/mistral-tiny":{"mode":"chat","base_model":"openrouter/mistralai/mistral-tiny","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_assistant_prefill":true,"supports_response_schema":true},"openrouter/mistralai/mistral-7b-instruct-v0.2":{"mode":"chat","base_model":"openrouter/mistralai/mistral-7b-instruct-v0.2","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_tool_choice":true},"openrouter/mistralai/mixtral-8x7b-instruct":{"mode":"chat","base_model":"openrouter/mistralai/mixtral-8x7b-instruct","provider":"litellm","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/neversleep/noromaid-20b":{"mode":"chat","base_model":"openrouter/neversleep/noromaid-20b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/alpindale/goliath-120b":{"mode":"chat","base_model":"openrouter/alpindale/goliath-120b","provider":"litellm","max_input_tokens":6144,"max_output_tokens":6144,"max_tokens":6144,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":6144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/undi95/toppy-m-7b":{"mode":"chat","base_model":"openrouter/undi95/toppy-m-7b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openrouter/auto":{"mode":"chat","base_model":"openrouter/openrouter/auto","provider":"litellm","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true},"openrouter/openai/gpt-4-1106-preview":{"mode":"chat","base_model":"openrouter/openai/gpt-4-1106-preview","provider":"litellm","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":65536,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_system_messages":true},"openrouter/mistralai/mistral-7b-instruct-v0.1":{"mode":"chat","base_model":"openrouter/mistralai/mistral-7b-instruct-v0.1","provider":"litellm","max_input_tokens":2824,"max_output_tokens":2824,"max_tokens":2824,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":1024,"range":{"min":1,"max":2824},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-3.5-turbo-instruct":{"mode":"chat","base_model":"openrouter/openai/gpt-3.5-turbo-instruct","provider":"litellm","max_input_tokens":4095,"max_output_tokens":4095,"max_tokens":4095,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4095},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/pygmalionai/mythalion-13b":{"mode":"chat","base_model":"openrouter/pygmalionai/mythalion-13b","provider":"litellm","max_input_tokens":4096,"max_output_tokens":4096,"max_tokens":4096,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"openrouter/openai/gpt-4-0314":{"mode":"chat","base_model":"openrouter/openai/gpt-4-0314","provider":"litellm","max_input_tokens":8191,"max_output_tokens":8191,"max_tokens":8191,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8191},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv"},"cerebras/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"cerebras/llama-4-scout-17b-16e-instruct","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"source":"merged_from_llm_models_csv","supports_vision":true,"supports_response_schema":false,"supports_reasoning":false},"cerebras/llama-4-maverick-17b-128e-instruct":{"mode":"chat","base_model":"cerebras/llama-4-maverick-17b-128e-instruct","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"Seed value for reproducibility","type":"number","default":0,"range":{"min":0,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"source":"merged_from_llm_models_csv"},"cerebras/qwen-3-235b-a22b-thinking-2507":{"mode":"chat","base_model":"cerebras/qwen-3-235b-a22b-thinking-2507","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning":true,"source":"merged_from_llm_models_csv"},"cerebras/qwen-3-coder-480b":{"mode":"chat","base_model":"cerebras/qwen-3-coder-480b","provider":"litellm","max_input_tokens":8192,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":2},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","default":50,"range":{"min":1,"max":100}},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"repetition_penalty","label":"Repetition Penalty","helpText":"A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.","type":"number","default":1,"range":{"min":1,"max":2}},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"merged_from_llm_models_csv"},"vertex_ai/claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","deprecation_date":"2027-04-16","regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":2048,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"vertex_multi_region_only":true,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/claude-opus-4-7@default":{"mode":"chat","base_model":"claude-opus-4-7","deprecation_date":"2027-04-16","regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":2048,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"vertex_multi_region_only":true,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/container-1g":{"mode":"chat","base_model":"container","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/container-4g":{"mode":"chat","base_model":"container","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/container-16g":{"mode":"chat","base_model":"container","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai/container-64g":{"mode":"chat","base_model":"container","provider":"openai","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1,"fast":6},"supports_output_config":true,"prompt_cache_min_tokens":2048,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","max_tokens":128000,"deprecation_date":"2027-04-16","inference_geo_us_multiplier":1.1,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"tool_use_system_prompt_tokens":346},"claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2027-02-17","supports_web_search":true,"inference_geo_us_multiplier":1.1,"supports_parallel_function_calling":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["auto","standard_only"],"supports_adaptive_thinking":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"tool_use_system_prompt_tokens":346,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"tool_use_system_prompt_tokens":346},"eu.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"eu-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"tool_use_system_prompt_tokens":346},"gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_web_search":true,"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true},"claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1,"fast":6},"supports_output_config":true,"supports_speed":true,"supports_fast_mode":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","max_tokens":1000000,"tool_use_system_prompt_tokens":346,"deprecation_date":"2027-05-28","inference_geo_us_multiplier":1.1,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true},"anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":1000000,"tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":1000000,"tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"anthropic/claude-opus-4.8":{"mode":"chat","base_model":"claude-opus-4-8","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"openrouter","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"thinking","label":"Thinking","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}]},{"id":"output_config","label":"output_config","type":"select","accesorKey":"effort","default":{"type":"text"},"options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}]}]},"claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://docs.anthropic.com/en/docs/about-claude/models/overview","provider":"anthropic","web_search_billing_unit":"per_request","max_tokens":1000000,"tool_use_system_prompt_tokens":346,"deprecation_date":"2027-06-09","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning":{"can_disable":false},"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":true,"supports_native_effort":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-fable-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":1000000,"tool_use_system_prompt_tokens":346,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"reasoning":{"can_disable":false},"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-5.4-2026-03-05":{"mode":"chat","base_model":"gpt-5.4","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_web_search":true,"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.4-mini-2026-03-17":{"mode":"chat","base_model":"gpt-5.4-mini","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"deprecation_date":"2027-04-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.4-nano-2026-03-17":{"mode":"chat","base_model":"gpt-5.4-nano","deprecation_date":"2027-04-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-luna","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://developers.openai.com/api/docs/pricing","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":-2,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"model-router":{"mode":"chat","base_model":"model-router","source":"https://azure.microsoft.com/en-us/pricing/details/ai-services/","comment":"Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/<deployment-name> where deployment-name is your Azure deployment (e.g., azure-model-router)","provider":"azure","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_reasoning_summary_detailed":false,"supports_reasoning_summary_concise":false,"supports_reasoning_summary_auto":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"openai.gpt-5.4":{"mode":"responses","base_model":"openai.gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_service_tier":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"supports_audio_input":false,"supports_audio_output":false},"openai.gpt-5.4-2026-03-05":{"mode":"responses","base_model":"openai.gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_service_tier":true,"provider":"bedrock_mantle","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openai.gpt-5.5":{"mode":"responses","base_model":"openai.gpt-5.5","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_web_search":true,"supports_reasoning_with_tool_calls":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_audio_input":false,"supports_audio_output":false},"openai.gpt-5.5-2026-04-23":{"mode":"responses","base_model":"openai.gpt-5.5","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_web_search":true,"supports_reasoning_with_tool_calls":false},"openai.gpt-5.6-luna":{"mode":"responses","base_model":"openai.gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"bedrock_mantle","supports_reasoning_with_tool_calls":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_audio_input":false,"supports_audio_output":false},"openai.gpt-5.6-terra":{"mode":"responses","base_model":"openai.gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider":"bedrock_mantle","supports_reasoning_with_tool_calls":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_audio_input":false,"supports_audio_output":false},"openai.gpt-5.6-sol":{"mode":"responses","base_model":"openai.gpt-5.6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider":"bedrock_mantle","supports_reasoning_with_tool_calls":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"response_format","type":"select","label":"Response Format","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","type":"json","label":"JSON Schema","default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response."}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","accesorKey":"type","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","type":"select","label":"Reasoning Effort","options":[{"label":"None","value":"none"},{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","type":"select","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_prompt_cache_breakpoint":true,"supports_audio_input":false,"supports_audio_output":false},"glm-5p2":{"mode":"responses","base_model":"glm-5.2","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses","/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_web_search":true},"glm-5.2":{"mode":"responses","base_model":"glm-5.2","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/responses","/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_web_search":true},"gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"supports_reasoning_with_tool_calls":false,"model":"gpt-5.6-sol","supports_custom_tools":true,"supports_service_tier":true,"service_tiers":["auto","default","flex","priority","fast"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true},"gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"supports_reasoning_with_tool_calls":false,"model":"gpt-5.6-luna","supports_service_tier":true,"service_tiers":["auto","default","flex","priority","fast"],"supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true},"gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"supports_reasoning_with_tool_calls":false,"model":"gpt-5.6-terra","supports_service_tier":true,"service_tiers":["auto","default","flex","priority","fast"],"supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true},"anthropic/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_output_config":true,"prompt_cache_min_tokens":1024,"provider":"anthropic","supports_web_search":true},"claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":1024,"source":"https://docs.anthropic.com/en/docs/about-claude/models/overview","provider":"anthropic","web_search_billing_unit":"per_request","max_tokens":128000,"deprecation_date":"2027-06-30","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_speed":true,"supports_fast_mode":true,"prompt_cache_min_tokens":512,"source":"https://docs.anthropic.com/en/docs/about-claude/models/overview","provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2027-07-24","tool_use_system_prompt_tokens":286,"supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true},"anthropic/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_speed":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"provider":"anthropic","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"tool_use_system_prompt_tokens":286,"supports_web_search":true},"gpt-5.3-codex":{"mode":"responses","base_model":"gpt-5.3-codex","max_input_tokens":272000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_contexts":["auto","current_turn"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"deprecation_date":"2027-04-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-6-sol","provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"model_parameters":[{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_service_tier":true,"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none"}},"gpt-5.5-2026-04-23":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_reasoning_with_tool_calls":false,"supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","flex","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_assistant_prefill":true},"bedrock_mantle/moonshotai.kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8192,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","ap-northeast-2","ap-southeast-1","eu-north-1","eu-west-2"],"is_deprecated":true,"max_tokens":8192,"supports_system_messages":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"bedrock_mantle/moonshotai.kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart","supports_audio_input":false,"supports_response_schema":true,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-north-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","ap-northeast-2","ap-southeast-1"],"max_tokens":262144,"supports_reasoning":true,"supports_video_input":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-3.1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","successor_model":"gemini-3.5-flash-lite","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini/gemini-3.1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_audio_input":true,"supports_audio_output":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.1-flash-lite","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_native_streaming":true,"supports_url_context":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/gemini-3.1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_audio_input":true,"supports_audio_output":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_video_length":1,"max_videos_per_prompt":10,"service_tiers":["priority","flex"],"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"reasoning_effort_levels":["minimal","low","medium","high"],"supports_multimodal_tool_output":true,"supports_native_streaming":true,"supports_response_schema_with_tools":true,"supports_video_input":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_web_search":true,"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true,"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"rpm":15,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"tpm":250000,"web_search_billing_unit":"per_query","provider":"gemini"},"gemini/gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-flash-lite","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":65536,"rpm":15,"tpm":250000,"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"vertex_ai/gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":65536,"service_tiers":["priority","flex"],"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"gemini-3.1-pro-preview":{"mode":"chat","base_model":"gemini-3.1-pro","max_input_tokens":1048576,"max_output_tokens":65536,"prompt_cache_min_tokens":4096,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":"media_resolution_medium","helpText":"Higher resolutions may provide better understanding but use more tokens.","id":"media_resolution","label":"Media Resolution","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini/gemini-3.1-pro-preview":{"mode":"chat","base_model":"gemini-3.1-pro","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65535,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.1-pro-preview","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_url_context":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"gemini","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"supports_audio_output":false,"supports_parallel_function_calling":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"thinking_level","label":"Thinking Level","helpText":"Set the thinking level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"Nucleus sampling probability mass.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"Text, JSON, or Structured output matching a supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema the model will follow.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"Number of most likely tokens to return at each position.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize tokens that appear in the text so far.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) best-effort deterministic sampling.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"Whether the response is sent incrementally or as one complete result.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/gemini-3.1-pro-preview":{"mode":"chat","base_model":"gemini-3.1-pro","max_input_tokens":1048576,"max_output_tokens":65535,"prompt_cache_min_tokens":4096,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","max_audio_length_hours":8.4,"max_audio_per_prompt":1,"max_images_per_prompt":3000,"max_pdf_size_mb":30,"max_tokens":65535,"max_video_length":1,"max_videos_per_prompt":10,"service_tiers":["priority","flex"],"supports_multimodal_tool_output":true,"supports_response_schema_with_tools":true,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call. Supports Google Search, File Search, Code Execution, URL Context, and Function Calling.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"media_resolution","label":"Media Resolution","helpText":"Higher resolutions may provide better understanding but use more tokens.","type":"select","default":"media_resolution_medium","options":[{"label":"Low","value":"media_resolution_low"},{"label":"Medium","value":"media_resolution_medium"},{"label":"High","value":"media_resolution_high"}]},{"id":"thinking_level","label":"Thinking Level","helpText":"Set the thinking level","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}]},{"id":"temperature","label":"Temperature","helpText":"For Gemini 3, best results at default 1.0. Lower values may impact reasoning.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Output length","helpText":"Maximum number of tokens in response","type":"number","default":32768,"range":{"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stopSequences","label":"Stop Sequences","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1}},{"id":"topP","label":"Top P","helpText":"Nucleus sampling probability mass.","type":"number","range":{"min":0,"max":1,"step":0.01}},{"id":"response_format","label":"Response Format","helpText":"Text, JSON, or Structured output matching a supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema the model will follow.","type":"json","default":{"type":"object","properties":{},"required":[],"propertyOrdering":[]}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Positive values penalize new tokens based on their existing frequency in the text so far.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"Number of most likely tokens to return at each position.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize tokens that appear in the text so far.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) best-effort deterministic sampling.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"Whether the response is sent incrementally or as one complete result.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"eu.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"eu-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"eu-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"eu.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"eu-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}],"supports_pdf_input":false,"supports_web_search":true},"DeepSeek-V4-Pro":{"mode":"chat","base_model":"DeepSeek-V4-Pro","max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"DeepSeek-V4-Flash":{"mode":"chat","base_model":"DeepSeek-V4-Flash","max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"DeepSeek-V3.2":{"mode":"chat","base_model":"DeepSeek-V3.2","max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"DeepSeek-V3.2-Speciale":{"mode":"chat","base_model":"DeepSeek-V3.2-Speciale","max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://platform.claude.com/docs/en/models/fable-5-1/whats-new-fable-5-1","provider":"anthropic","web_search_billing_unit":"per_request","supports_forced_tool_choice":false,"inference_geo_us_multiplier":1.1,"tool_use_system_prompt_tokens":346,"supports_safeguards":true,"deprecation_date":"2027-09-01","max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning":{"can_disable":false},"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","standard_only"],"supports_dynamic_reasoning_budget":false,"supports_mid_conversation_system_messages":true,"supports_native_effort":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":1000000,"supports_forced_tool_choice":false,"supports_web_search":true,"tool_use_system_prompt_tokens":346,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"reasoning":{"can_disable":false},"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"reasoning":{"can_disable":false},"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"supports_forced_tool_choice":false,"supports_web_search":true,"tool_use_system_prompt_tokens":346,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"reasoning":{"can_disable":false},"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"eu.anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_forced_tool_choice":false,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_native_structured_output":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_web_search":true,"prompt_cache_min_tokens":512,"provider":"bedrock","reasoning":{"can_disable":false},"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"vertex_ai/claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-03-01","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"reasoning":{"can_disable":false},"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-fable-5-1@default":{"mode":"chat","base_model":"claude-fable-5-1","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-03-01","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"reasoning":{"can_disable":false},"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-12-05","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","provider":"azure","supports_forced_tool_choice":false,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"reasoning":{"can_disable":false},"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"global.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"global.anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_parallel_function_calling":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"global.anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"supports_parallel_function_calling":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"anthropic.claude-mythos-preview":{"mode":"chat","base_model":"claude-mythos","max_input_tokens":1000000,"max_output_tokens":128000,"thinking_always_on":true,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_output_config":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"au.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"us.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_parallel_function_calling":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"us.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"au.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"au.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"au.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"au.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"us.anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-fable-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1}},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":0.7,"range":{"min":0,"max":1},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_k","label":"Top K","helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","type":"number","range":{"min":0,"max":100}},{"id":"output_format","label":"Output Format","helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","options":[{"label":"Output output","value":"json_schema","subFields":[{"id":"schema","label":"Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"type":"object"}}]}]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"thinking","label":"Thinking","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}]}]},"bedrock/us-gov-east-1/anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"global.anthropic.claude-fable-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"global.anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"global.anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"global.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_parallel_function_calling":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_adaptive_thinking":true,"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"jp.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"ap-northeast-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false},"jp.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"ap-northeast-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_service_tier":true,"supports_speed":false,"supports_web_search":true},"jp.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"jp.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_max_reasoning_effort":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-6","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["low","medium","high","max"],"service_tiers":["default"],"supports_cache_point":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":true,"supports_speed":false},"jp.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"bedrock/us-gov-east-1/anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-27","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"us-gov.anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/us-gov.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"provider":"bedrock","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2027-04-08","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"bedrock/us-gov-west-1/anthropic.claude-sonnet-5-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":8192,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-4-5-20250929-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"max_tokens":8192,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":8192,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/us-gov.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"provider":"bedrock","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"eu.anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":16384,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","deprecation_date":"2028-01-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6-astra-2026-09-03":{"mode":"chat","base_model":"gpt-6-astra","deprecation_date":"2028-01-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","deprecation_date":"2028-01-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-astra-2026-09-03":{"mode":"chat","base_model":"gpt-6-astra","max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6-astra","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"supports_service_tier":true,"service_tiers":["auto","default","flex","priority","fast","ultrafast"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true},"gemini-3.1-pro-preview-customtools":{"mode":"chat","base_model":"gemini-3.1-pro-customtools","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_url_context":true,"supports_native_streaming":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini-3.1-flash-lite-preview":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_web_search":true,"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true,"supports_audio_output":false,"supports_audio_input":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_native_streaming":true,"supports_code_execution":true,"supports_file_search":true,"max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"web_search_billing_unit":"per_query","provider":"gemini"},"gemini-3.6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"deprecation_date":"2026-11-19","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"gemini-3.8-flash","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"deprecation_date":"2027-01-28","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"gemini-3.8-flash","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini-3.8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","max_input_tokens":1048576,"max_output_tokens":65536,"prompt_cache_min_tokens":4096,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":false,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"gemini-robotics-er-2-preview":{"mode":"chat","base_model":"gemini-robotics-er-2","max_input_tokens":131072,"max_output_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"high","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Minimal","value":"minimal"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":65536},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_service_tier":true,"service_tiers":["default","flex","priority"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["minimal","low","medium","high"],"supports_reasoning_disable":true,"supports_none_reasoning_effort":true,"supports_dynamic_reasoning_budget":true,"reasoning_budget":{"min":1,"max":65535},"supports_response_schema_with_tools":true},"chat-latest":{"mode":"chat","base_model":"chat","max_input_tokens":400000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"reasoning_effort_levels":["medium"],"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard"],"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"supports_service_tier":true,"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":false,"supports_reasoning_with_tool_calls":false},"codex-auto-review":{"mode":"chat","base_model":"gpt-5.4","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_service_tier":true,"supports_vision":true,"provider":"openai","supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["none","low","medium","high","xhigh"],"supports_web_search":true,"supports_assistant_prefill":true,"supports_none_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_reasoning_with_tool_calls":false},"gpt-5.4-pro":{"mode":"responses","base_model":"gpt-5.4-pro","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"]},"gpt-5.4-pro-2026-03-05":{"mode":"responses","base_model":"gpt-5.4-pro","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"]},"gpt-5.5-pro":{"mode":"responses","base_model":"gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"]},"gpt-5.5-pro-2026-04-23":{"mode":"responses","base_model":"gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":false,"provider":"openai","web_search_billing_unit":"per_request","supports_service_tier":true,"supports_custom_tools":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","priority"],"reasoning_effort_levels":["medium","high","xhigh"]},"gpt-5.6":{"mode":"chat","base_model":"gpt-5.6","max_input_tokens":922000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","service_tiers":["auto","flex","priority"],"supports_custom_tools":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_service_tier":true,"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":false},"claude-haiku-4-5@20251001":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2026-10-15","max_input_tokens":200000,"max_output_tokens":8192,"max_tokens":8192,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":8192,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":4096,"provider":"vertex_ai","regional_endpoint_uplift_multiplier":1.1,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/haiku-4-5","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"is_deprecated":true},"claude-opus-4":{"mode":"chat","base_model":"claude-opus-4","deprecation_date":"2026-05-14","is_deprecated":true,"max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":32000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":159},"claude-opus-4-1@20250805":{"mode":"chat","base_model":"claude-opus-4-1","deprecation_date":"2026-08-05","is_deprecated":true,"max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":32000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"vertex_ai","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true},"claude-opus-4-5@20251101":{"mode":"chat","base_model":"claude-opus-4-5","deprecation_date":"2026-11-24","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":4096,"provider":"vertex_ai-anthropic_models","reasoning_effort_levels":["low","medium","high"],"regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_native_streaming":true,"supports_output_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":159},"claude-opus-4-6@default":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","max"],"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":346},"claude-opus-4-7@default":{"mode":"chat","base_model":"claude-opus-4-7","deprecation_date":"2027-04-16","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":2048,"provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","xhigh","max"],"regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system_messages":false,"supports_minimal_reasoning_effort":true,"supports_native_effort":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"tool_use_system_prompt_tokens":346,"vertex_multi_region_only":true,"supports_safeguards":true},"claude-opus-4-8@default":{"mode":"chat","base_model":"claude-opus-4-8","deprecation_date":"2027-05-28","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","xhigh","max"],"regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"supports_safeguards":true},"claude-opus-4@20250514":{"mode":"chat","base_model":"claude-opus-4","deprecation_date":"2026-05-14","is_deprecated":true,"max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":32000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":159},"claude-sonnet-4":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-05-14","is_deprecated":true,"max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":159},"claude-sonnet-4-5@20250929":{"mode":"chat","base_model":"claude-sonnet-4-5","deprecation_date":"2026-09-29","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"is_deprecated":true},"claude-sonnet-4-6@default":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","max"],"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":346},"claude-sonnet-4@20250514":{"mode":"chat","base_model":"claude-sonnet-4","deprecation_date":"2026-05-14","is_deprecated":true,"max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","supports_adaptive_thinking":false,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":true,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"tool_use_system_prompt_tokens":159},"claude-sonnet-5@default":{"mode":"chat","base_model":"claude-sonnet-5","deprecation_date":"2026-12-24","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"prompt_cache_min_tokens":1024,"provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","xhigh","max"],"regional_endpoint_uplift_multiplier":1.1,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_dynamic_reasoning_budget":false,"supports_forced_tool_choice":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_native_effort":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":true,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_speed":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"supports_safeguards":true},"bedrock_mantle/xai.grok-4.3":{"mode":"responses","base_model":"grok-4.3","use_openai_responses_path":true,"max_input_tokens":131072,"max_output_tokens":16384,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-east-1","us-east-2","us-gov-west-1"],"max_tokens":16384,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Modify the likelihood of specified tokens appearing in the completion.","id":"logit_bias","label":"Logit Bias","range":{"max":100,"min":-100},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"A unique identifier representing your end-user.","id":"metadata","label":"Metadata","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":true,"supports_web_search":false,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":false,"top_p":false,"web_search_options":false}},"bedrock_mantle/xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","use_openai_responses_path":true,"max_input_tokens":500000,"max_output_tokens":500000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-gov-east-1","us-east-1","us-east-2","us-west-1","us-gov-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Modify the likelihood of specified tokens appearing in the completion.","id":"logit_bias","label":"Logit Bias","range":{"max":100,"min":-100},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"A unique identifier representing your end-user.","id":"metadata","label":"Metadata","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":true,"supports_web_search":false,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":false,"top_p":false,"web_search_options":false}},"global.xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":500000,"max_output_tokens":500000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"unsupported_fields":{"stop":true,"top_p":true}},"grok-3":{"mode":"chat","base_model":"grok-3","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-3-beta":{"mode":"chat","base_model":"grok-3","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-fast-beta":{"mode":"chat","base_model":"grok-3-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-fast-latest":{"mode":"chat","base_model":"grok-3-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-latest":{"mode":"chat","base_model":"grok-3","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-mini":{"mode":"chat","base_model":"grok-3-mini","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-3-mini-beta":{"mode":"chat","base_model":"grok-3-mini","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-mini-fast":{"mode":"chat","base_model":"grok-3-mini-fast","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-mini-fast-beta":{"mode":"chat","base_model":"grok-3-mini-fast","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-mini-fast-latest":{"mode":"chat","base_model":"grok-3-mini-fast","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-3-mini-latest":{"mode":"chat","base_model":"grok-3-mini","deprecation_date":"2026-02-28","is_deprecated":true,"max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://x.ai/api#pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4":{"mode":"chat","base_model":"grok-4","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-4-0709":{"mode":"chat","base_model":"grok-4","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4-1-fast":{"mode":"chat","base_model":"grok-4-1-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_assistant_prefill":true,"supports_audio_input":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4-1-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-1-fast-non","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning","supports_assistant_prefill":true,"supports_audio_input":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4-1-fast-non-reasoning-latest":{"mode":"chat","base_model":"grok-4-1-fast-non","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning","supports_assistant_prefill":true,"supports_audio_input":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4-1-fast-reasoning":{"mode":"chat","base_model":"grok-4-1-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_assistant_prefill":true,"supports_audio_input":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-4-1-fast-reasoning-latest":{"mode":"chat","base_model":"grok-4-1-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_assistant_prefill":true,"supports_audio_input":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-fast-non","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-4-fast-reasoning":{"mode":"chat","base_model":"grok-4-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_pdf_input":false},"grok-4-latest":{"mode":"chat","base_model":"grok-4","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"grok-4.20":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-0309":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-0309-non-reasoning":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-0309-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-0309":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-0309-non-reasoning":{"mode":"chat","base_model":"grok-4.20-beta-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-0309-reasoning":{"mode":"chat","base_model":"grok-4.20-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-latest":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-latest-non-reasoning":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-latest-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-non-reasoning":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-beta-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-0304":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-0304-non-reasoning":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-0304-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-latest":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-non-reasoning-latest":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-experimental-beta-reasoning-latest":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-0309":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-beta-0309":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-beta-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-experimental-beta-0304":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-experimental-beta-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-multi-agent-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"top_p":false},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-non-reasoning":{"mode":"chat","base_model":"grok-4.20-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-non-reasoning-gv2":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-non-reasoning-latest":{"mode":"chat","base_model":"grok-4.20-non","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-reasoning-gv2":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.20-reasoning-latest":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.3":{"mode":"chat","base_model":"grok-4.3","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.3-latest":{"mode":"chat","base_model":"grok-4.3","max_input_tokens":1000000,"max_output_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.5":{"mode":"chat","base_model":"grok-4.5","max_input_tokens":500000,"max_output_tokens":500000,"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","reasoning_effort_levels":["minimal","low","medium","high"],"source":"https://docs.x.ai/developers/models/grok-4.5","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.5-latest":{"mode":"chat","base_model":"grok-4.5","max_input_tokens":500000,"max_output_tokens":500000,"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","reasoning_effort_levels":["minimal","low","medium","high"],"source":"https://docs.x.ai/developers/models/grok-4.5","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":500000,"max_output_tokens":500000,"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","reasoning_effort_levels":["minimal","low","medium","high","xhigh"],"source":"https://docs.x.ai/developers/models/grok-4.6","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-build-0.1":{"mode":"chat","base_model":"grok-build-0.1","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/developers/models/grok-build-0.1","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-build-latest":{"mode":"chat","base_model":"grok-build","max_input_tokens":500000,"max_output_tokens":500000,"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","reasoning_effort_levels":["minimal","low","medium","high"],"source":"https://docs.x.ai/developers/models","supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-code-fast":{"mode":"chat","base_model":"grok-code-fast","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-code-fast-1":{"mode":"chat","base_model":"grok-code-fast-1","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"grok-code-fast-1-0825":{"mode":"chat","base_model":"grok-code-fast-1","deprecation_date":"2026-05-15","is_deprecated":true,"max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"xai","source":"https://docs.x.ai/docs/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"reasoning_effort":true,"stop":true,"top_p":false,"web_search_options":true},"supports_audio_input":false,"supports_audio_output":false},"us.xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":500000,"max_output_tokens":500000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":true,"unsupported_fields":{"stop":true,"top_p":true}},"vertex_ai/xai/grok-4.1-fast-non-reasoning":{"mode":"chat","base_model":"grok-4.1-fast-non-reasoning","max_input_tokens":2000000,"max_output_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","max_tokens":2000000,"supports_audio_input":true,"supports_reasoning":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","id":"frequency_penalty","label":"Frequency Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":0,"helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","id":"presence_penalty","label":"Presence Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_service_tier":false,"unsupported_fields":{"frequency_penalty":false,"presence_penalty":false,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true}},"vertex_ai/xai/grok-4.1-fast-reasoning":{"mode":"chat","base_model":"grok-4.1-fast-non-reasoning","max_input_tokens":2000000,"max_output_tokens":2000000,"source":"https://docs.x.ai/docs/models/grok-4-1-fast-reasoning","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","max_tokens":2000000,"supports_audio_input":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","id":"frequency_penalty","label":"Frequency Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":0,"helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","id":"presence_penalty","label":"Presence Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":false,"unsupported_fields":{"frequency_penalty":false,"presence_penalty":false,"stop":false,"top_p":false,"web_search_options":true}},"vertex_ai/xai/grok-4.20-non-reasoning":{"mode":"chat","base_model":"grok-4.20-non","max_input_tokens":2000000,"max_output_tokens":2000000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"vertex_ai","max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","id":"frequency_penalty","label":"Frequency Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":0,"helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","id":"presence_penalty","label":"Presence Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":false,"supports_service_tier":false,"unsupported_fields":{"frequency_penalty":false,"presence_penalty":false,"reasoning_effort":true,"stop":false,"top_p":false,"web_search_options":true}},"vertex_ai/xai/grok-4.20-reasoning":{"mode":"chat","base_model":"grok-4.20-reasoning","max_input_tokens":2000000,"max_output_tokens":2000000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","max_tokens":2000000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","id":"frequency_penalty","label":"Frequency Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":0,"helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","id":"presence_penalty","label":"Presence Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":false,"unsupported_fields":{"frequency_penalty":false,"presence_penalty":false,"stop":false,"top_p":false,"web_search_options":true}},"vertex_ai/xai/grok-4.3":{"mode":"chat","base_model":"grok-4.3","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://docs.x.ai/developers/models/grok-4.3","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":1000000,"supports_web_search":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","id":"frequency_penalty","label":"Frequency Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"A unique identifier representing your end-user.","id":"metadata","label":"Metadata","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":0,"helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","id":"presence_penalty","label":"Presence Penalty","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":true,"unsupported_fields":{"frequency_penalty":false,"presence_penalty":false,"stop":false,"top_p":false,"web_search_options":true}},"vertex_ai/xai/grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":524288,"max_output_tokens":524288,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":524288,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","id":"top_logprobs","label":"Top Logprobs","range":{"max":5,"min":0,"step":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_service_tier":true,"supports_web_search":false,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":true,"top_p":false,"web_search_options":true}},"xai.grok-4.3":{"mode":"responses","base_model":"grok-4.3","max_input_tokens":131072,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Modify the likelihood of specified tokens appearing in the completion.","id":"logit_bias","label":"Logit Bias","range":{"max":100,"min":-100},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"A unique identifier representing your end-user.","id":"metadata","label":"Metadata","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":false,"top_p":false,"web_search_options":false},"use_openai_responses_path":true,"supports_audio_input":false,"supports_audio_output":false},"xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":500000,"max_output_tokens":500000,"max_tokens":500000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Modify the likelihood of specified tokens appearing in the completion.","id":"logit_bias","label":"Logit Bias","range":{"max":100,"min":-100},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":false,"helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","id":"logprobs","label":"Logprobs","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"helpText":"A unique identifier representing your end-user.","id":"metadata","label":"Metadata","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","id":"seed","label":"Seed","range":{"max":100,"min":0},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":2,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"default":65536,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_completion_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions"]},{"default":8192,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_output_tokens","label":"Max Tokens","type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","id":"max_tool_calls","label":"Max Tool Calls","range":{"max":128,"min":1,"step":1},"type":"number","supported_endpoints":["/v1/responses"]},{"helpText":"Static predicted output content, such as the contents of a text file being regenerated with minor changes.","id":"prediction","label":"Predicted Output","type":"text","supported_endpoints":["/v1/chat/completions"]},{"helpText":"Used to cache responses for similar requests to optimize your cache hit rates.","id":"prompt_cache_key","label":"Prompt Cache Key","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"The retention policy for the prompt cache.","id":"prompt_cache_retention","label":"Prompt Cache Retention","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","id":"verbosity","label":"Verbosity","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Whether to stream back partial progress.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"provider":"bedrock_mantle","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_service_tier":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"unsupported_fields":{"frequency_penalty":true,"presence_penalty":true,"stop":false,"top_p":false,"web_search_options":false},"use_openai_responses_path":true,"supports_audio_input":false,"supports_audio_output":false},"amazon.titan-embed-g1-text-02":{"mode":"embedding","base_model":"titan-embed-g1-text-02","max_input_tokens":8192,"output_vector_size":1536,"provider":"bedrock","max_tokens":8192},"azure/us-gov/text-embedding-3-large":{"mode":"embedding","base_model":"us-gov/text-embedding-3-large","max_input_tokens":8191,"provider":"azure","max_tokens":8191},"azure/us-gov/text-embedding-3-small":{"mode":"embedding","base_model":"us-gov/text-embedding-3-small","max_input_tokens":8191,"provider":"azure","max_tokens":8191},"bedrock/us-gov-west-1/amazon.nova-2-multimodal-embeddings-v1:0":{"mode":"embedding","base_model":"nova-2-multimodal-embeddings","max_input_tokens":8172,"output_vector_size":3072,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_embedding_image_input":true,"supports_image_input":true,"supports_video_input":true,"supports_audio_input":true,"provider":"bedrock","max_tokens":8172},"cloudflare/workers-ai/@cf/baai/bge-base-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-base-en-v1.5","max_input_tokens":153600,"max_tokens":153600,"provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"cloudflare/workers-ai/@cf/baai/bge-large-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-large-en-v1.5","provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"cloudflare/workers-ai/@cf/baai/bge-m3":{"mode":"embedding","base_model":"@cf/baai/bge-m3","max_input_tokens":60000,"max_tokens":60000,"provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"cloudflare/workers-ai/@cf/baai/bge-small-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-small-en-v1.5","provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"cloudflare/workers-ai/@cf/pfnet/plamo-embedding-1b":{"mode":"embedding","base_model":"@cf/pfnet/plamo-embedding-1b","provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"cloudflare/workers-ai/@cf/qwen/qwen3-embedding-0.6b":{"mode":"embedding","base_model":"@cf/qwen/qwen3-embedding-0.6b","max_input_tokens":8192,"max_tokens":8192,"provider":"cloudflare","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"databricks/databricks-qwen3-embedding-0-6b":{"mode":"embedding","base_model":"qwen3-embedding-0.6b","max_input_tokens":32768,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Public Preview. Configurable output dimensionality up to 1024."},"output_vector_size":1024,"source":"https://www.databricks.com/product/pricing/foundation-model-serving","provider":"databricks","max_tokens":32768},"databricks/gte-large-en":{"mode":"embedding","base_model":"databricks-gte-large-en","provider":"databricks","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"databricks/qwen3-embedding-0-6b":{"mode":"embedding","base_model":"databricks-qwen3-embedding-0-6b","max_input_tokens":8192,"max_tokens":8192,"provider":"databricks","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"databricks/system.ai.gte-large-en":{"mode":"embedding","base_model":"databricks-gte-large-en","provider":"databricks","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"databricks/system.ai.qwen3-embedding-0-6b":{"mode":"embedding","base_model":"databricks-qwen3-embedding-0-6b","provider":"databricks","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":8192,"max_tokens":8192,"max_output_tokens":32768},"eu.twelvelabs.marengo-embed-3-0-v1:0":{"mode":"embedding","base_model":"marengo-embed-3-0","max_input_tokens":500,"output_vector_size":512,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":500},"gemini-embedding-2":{"mode":"embedding","base_model":"gemini-embedding-2","max_input_tokens":8192,"output_vector_size":3072,"source":"https://ai.google.dev/gemini-api/docs/pricing","supports_embedding_image_input":true,"supports_multimodal":true,"uses_embed_content":true,"provider":"gemini","max_tokens":8192,"rpm":10000,"tpm":10000000},"gemini/gemini-embedding-2":{"mode":"embedding","base_model":"gemini-embedding-2","max_input_tokens":8192,"output_vector_size":3072,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supports_audio_input":true,"supports_embedding_image_input":true,"supports_multimodal":true,"supports_vision":true,"tpm":10000000,"provider":"gemini","max_tokens":8192},"gemini/gemini-embedding-2-preview":{"mode":"embedding","base_model":"gemini-embedding-2","deprecation_date":"2026-08-10","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","max_input_tokens":8192,"output_vector_size":3072,"rpm":10000,"source":"https://ai.google.dev/gemini-api/docs/pricing","successor_model":"gemini/gemini-embedding-2","supports_audio_input":true,"supports_embedding_image_input":true,"supports_multimodal":true,"supports_vision":true,"tpm":10000000,"provider":"gemini","is_deprecated":true,"max_tokens":8192},"gigachat/GigaEmbeddings-3B-2025-09":{"mode":"embedding","base_model":"gigaembeddings-3b-2025-09","max_input_tokens":4096,"output_vector_size":2048,"provider":"gigachat","max_tokens":4096},"global.cohere.embed-v4:0":{"mode":"embedding","base_model":"embed","max_input_tokens":128000,"output_vector_size":1536,"supports_embedding_image_input":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","sa-east-1"],"max_tokens":128000},"libertai/bge-m3":{"mode":"embedding","base_model":"bge-m3","max_input_tokens":8192,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":8192},"mistral/mistral-embed-2312":{"mode":"embedding","base_model":"mistral-embed","max_input_tokens":8192,"source":"https://docs.mistral.ai/models/mistral-embed-23-12","provider":"mistral","max_tokens":8192},"nebius/BAAI/bge-en-icl":{"mode":"embedding","base_model":"bge-en-icl","max_input_tokens":32768,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768},"nebius/BAAI/bge-multilingual-gemma2":{"mode":"embedding","base_model":"bge-multilingual-gemma2","max_input_tokens":8192,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":8192},"nebius/Qwen/Qwen3-Embedding-8B":{"mode":"embedding","base_model":"qwen3-embedding-8b","max_input_tokens":40960,"source":"https://tokenfactory.nebius.com/models/catalog/embedding/Qwen%2FQwen3-Embedding-8B","provider":"nebius","max_tokens":40960},"nebius/intfloat/e5-mistral-7b-instruct":{"mode":"embedding","base_model":"intfloat/e5-mistral-7b-instruct","max_input_tokens":32768,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768},"oci/cohere.embed-english-image-v3.0":{"mode":"embedding","base_model":"embed-english-image","max_input_tokens":512,"output_vector_size":1024,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_embedding_image_input":true,"provider":"oci","max_tokens":512},"oci/cohere.embed-english-light-image-v3.0":{"mode":"embedding","base_model":"embed-english-light-image","max_input_tokens":512,"output_vector_size":384,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_embedding_image_input":true,"provider":"oci","max_tokens":512},"oci/cohere.embed-english-light-v3.0":{"mode":"embedding","base_model":"embed-english-light","max_input_tokens":512,"output_vector_size":384,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","provider":"oci","max_tokens":512},"oci/cohere.embed-english-v3.0":{"mode":"embedding","base_model":"embed-english","max_input_tokens":512,"output_vector_size":1024,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","provider":"oci","max_tokens":512},"oci/cohere.embed-multilingual-image-v3.0":{"mode":"embedding","base_model":"embed-multilingual-image","max_input_tokens":512,"output_vector_size":1024,"source":"https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/","supports_embedding_image_input":true,"supports_vision":true,"provider":"oci"},"oci/cohere.embed-multilingual-light-image-v3.0":{"mode":"embedding","base_model":"embed-multilingual-light-image","max_input_tokens":512,"output_vector_size":384,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_embedding_image_input":true,"provider":"oci","max_tokens":512},"oci/cohere.embed-multilingual-light-v3.0":{"mode":"embedding","base_model":"embed-multilingual-light","max_input_tokens":512,"output_vector_size":384,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","provider":"oci","max_tokens":512},"oci/cohere.embed-multilingual-v3.0":{"mode":"embedding","base_model":"embed-multilingual","max_input_tokens":512,"output_vector_size":1024,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","provider":"oci","max_tokens":512},"oci/cohere.embed-v4.0":{"mode":"embedding","base_model":"embed","max_input_tokens":128000,"output_vector_size":1536,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_embedding_image_input":true,"provider":"oci","max_tokens":128000},"openrouter/baai/bge-base-en-v1.5":{"mode":"embedding","base_model":"bge-base-en-v1.5","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/baai/bge-large-en-v1.5":{"mode":"embedding","base_model":"bge-large-en-v1.5","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/baai/bge-m3":{"mode":"embedding","base_model":"bge-m3","provider":"openrouter","max_input_tokens":8194,"max_output_tokens":7374,"max_tokens":7374,"supports_response_schema":true},"openrouter/google/gemini-embedding-001":{"mode":"embedding","base_model":"gemini-embedding-001","provider":"openrouter","max_input_tokens":20000,"max_output_tokens":18000,"max_tokens":18000,"supports_response_schema":true},"openrouter/google/gemini-embedding-2":{"mode":"embedding","base_model":"gemini-embedding-2","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_response_schema":true},"openrouter/google/gemini-embedding-2-preview":{"mode":"embedding","base_model":"gemini-embedding-2-preview","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_response_schema":true},"openrouter/google/gemini-embedding-2:batch":{"mode":"embedding","base_model":"gemini-embedding-2:batch","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_response_schema":true},"openrouter/intfloat/e5-base-v2":{"mode":"embedding","base_model":"e5-base-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/intfloat/e5-large-v2":{"mode":"embedding","base_model":"e5-large-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/intfloat/multilingual-e5-large":{"mode":"embedding","base_model":"multilingual-e5-large","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/mistralai/codestral-embed-2505":{"mode":"embedding","base_model":"codestral-embed-2505","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":6553,"max_tokens":6553,"supports_response_schema":true},"openrouter/mistralai/mistral-embed-2312":{"mode":"embedding","base_model":"mistral-embed-2312","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":6553,"max_tokens":6553,"supports_response_schema":true},"openrouter/openai/text-embedding-3-large":{"mode":"embedding","base_model":"text-embedding-3-large","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/openai/text-embedding-3-large:batch":{"mode":"embedding","base_model":"text-embedding-3-large:batch","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/openai/text-embedding-3-small":{"mode":"embedding","base_model":"text-embedding-3-small","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/openai/text-embedding-3-small:batch":{"mode":"embedding","base_model":"text-embedding-3-small:batch","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/openai/text-embedding-ada-002":{"mode":"embedding","base_model":"text-embedding-ada-002","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/openai/text-embedding-ada-002:batch":{"mode":"embedding","base_model":"text-embedding-ada-002:batch","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_response_schema":true},"openrouter/perplexity/pplx-embed-v1-0.6b":{"mode":"embedding","base_model":"pplx-embed-v1-0.6b","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/perplexity/pplx-embed-v1-4b":{"mode":"embedding","base_model":"pplx-embed-v1-4b","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/qwen/qwen3-embedding-4b":{"mode":"embedding","base_model":"qwen3-embedding-4b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_response_schema":true},"openrouter/qwen/qwen3-embedding-8b":{"mode":"embedding","base_model":"qwen3-embedding-8b","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_response_schema":true},"openrouter/sentence-transformers/all-minilm-l12-v2":{"mode":"embedding","base_model":"all-minilm-l12-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/sentence-transformers/all-minilm-l6-v2":{"mode":"embedding","base_model":"all-minilm-l6-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/sentence-transformers/all-mpnet-base-v2":{"mode":"embedding","base_model":"all-mpnet-base-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/sentence-transformers/multi-qa-mpnet-base-dot-v1":{"mode":"embedding","base_model":"multi-qa-mpnet-base-dot-v1","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/sentence-transformers/paraphrase-minilm-l6-v2":{"mode":"embedding","base_model":"paraphrase-minilm-l6-v2","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/thenlper/gte-base":{"mode":"embedding","base_model":"gte-base","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/thenlper/gte-large":{"mode":"embedding","base_model":"gte-large","provider":"openrouter","max_input_tokens":512,"max_output_tokens":460,"max_tokens":460,"supports_response_schema":true},"openrouter/voyageai/voyage-4":{"mode":"embedding","base_model":"voyage-4","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"provider":"openrouter"},"openrouter/voyageai/voyage-4-large":{"mode":"embedding","base_model":"voyage-4-large","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"provider":"openrouter"},"openrouter/voyageai/voyage-4-lite":{"mode":"embedding","base_model":"voyage-4-lite","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"provider":"openrouter"},"openrouter/voyageai/voyage-code-4":{"mode":"embedding","base_model":"voyage-code-4","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"provider":"openrouter"},"openrouter/voyageai/voyage-multimodal-3.5":{"mode":"embedding","base_model":"voyage-multimodal-3.5","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"provider":"openrouter","supports_vision":true,"supports_embedding_image_input":true},"perplexity/pplx-embed-context-v1-0.6b":{"mode":"embedding","base_model":"pplx-embed-context-v1-0.6b","max_input_tokens":32768,"output_vector_size":1024,"source":"https://docs.perplexity.ai/getting-started/pricing","provider":"perplexity","max_tokens":32768},"perplexity/pplx-embed-context-v1-4b":{"mode":"embedding","base_model":"pplx-embed-context-v1-4b","max_input_tokens":32768,"output_vector_size":2560,"source":"https://docs.perplexity.ai/getting-started/pricing","provider":"perplexity","max_tokens":32768},"perplexity/pplx-embed-v1-0.6b":{"mode":"embedding","base_model":"pplx-embed-v1-0.6b","max_input_tokens":32768,"output_vector_size":1024,"source":"https://docs.perplexity.ai/docs/embeddings/quickstart","provider":"perplexity","max_tokens":32768},"perplexity/pplx-embed-v1-4b":{"mode":"embedding","base_model":"pplx-embed-v1-4b","max_input_tokens":32768,"output_vector_size":2560,"source":"https://docs.perplexity.ai/docs/embeddings/quickstart","provider":"perplexity","max_tokens":32768},"scaleway/BAAI/bge-multilingual-gemma2":{"mode":"embedding","base_model":"bge-multilingual-gemma2","provider":"scaleway"},"scaleway/qwen/qwen3-embedding-8b":{"mode":"embedding","base_model":"qwen3-embedding-8b","provider":"scaleway"},"snowflake/snowflake-arctic-embed-l-v2.0":{"mode":"embedding","base_model":"snowflake-arctic-embed-l","max_input_tokens":8192,"provider":"snowflake","max_tokens":8192},"snowflake/snowflake-arctic-embed-m-v2.0":{"mode":"embedding","base_model":"snowflake-arctic-embed-m","max_input_tokens":8192,"provider":"snowflake","max_tokens":8192},"together_ai/intfloat/multilingual-e5-large-instruct":{"mode":"embedding","base_model":"intfloat/multilingual-e5-large-instruct","max_input_tokens":514,"max_tokens":514,"output_vector_size":1024,"provider":"together_ai","source":"https://docs.together.ai/docs/serverless-models"},"twelvelabs.marengo-embed-3-0-v1:0":{"mode":"embedding","base_model":"marengo-embed-3-0","max_input_tokens":500,"output_vector_size":512,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":500},"us.cohere.embed-v4:0":{"mode":"embedding","base_model":"embed","max_input_tokens":128000,"output_vector_size":1536,"supports_embedding_image_input":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","sa-east-1"],"max_tokens":128000},"us.twelvelabs.marengo-embed-3-0-v1:0":{"mode":"embedding","base_model":"marengo-embed-3-0","max_input_tokens":500,"output_vector_size":512,"supports_embedding_image_input":true,"supports_image_input":true,"provider":"bedrock","max_tokens":500},"vertex_ai/gemini-embedding-2":{"mode":"embedding","base_model":"gemini-embedding-2","max_input_tokens":8192,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supports_embedding_image_input":true,"supports_multimodal":true,"uses_embed_content":true,"provider":"vertex_ai","max_tokens":8192},"vertex_ai/gemini-embedding-2-preview":{"mode":"embedding","base_model":"gemini-embedding-2","max_input_tokens":8192,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supports_embedding_image_input":true,"supports_multimodal":true,"uses_embed_content":true,"provider":"vertex_ai","max_tokens":8192},"vertex_ai/gemini-flash-experimental":{"mode":"embedding","base_model":"gemini-flash","max_input_tokens":1000000,"max_output_tokens":8192,"max_tokens":8192,"output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","uses_embed_content":true,"provider":"vertex_ai"},"voyage/voyage-4":{"mode":"embedding","base_model":"voyage-4","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":32000},"voyage/voyage-4-large":{"mode":"embedding","base_model":"voyage-4-large","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":32000},"voyage/voyage-4-lite":{"mode":"embedding","base_model":"voyage-4-lite","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":32000},"voyage/voyage-code-4":{"mode":"embedding","base_model":"voyage-code-4","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":32000},"voyage/voyage-context-4":{"mode":"embedding","base_model":"voyage-context-4","max_input_tokens":120000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":120000},"voyage/voyage-multilingual-2":{"mode":"embedding","base_model":"voyage-multilingual-2","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","max_tokens":32000},"voyage/voyage-multimodal-3.5":{"mode":"embedding","base_model":"voyage-multimodal-3.5","max_input_tokens":32000,"output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","supports_embedding_image_input":true,"provider":"voyage","max_tokens":32000},"workers-ai/workers-ai/@cf/baai/bge-base-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-base-en-v1.5","max_input_tokens":153600,"max_tokens":153600,"provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"workers-ai/workers-ai/@cf/baai/bge-large-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-large-en-v1.5","provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"workers-ai/workers-ai/@cf/baai/bge-m3":{"mode":"embedding","base_model":"@cf/baai/bge-m3","max_input_tokens":60000,"max_tokens":60000,"provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"workers-ai/workers-ai/@cf/baai/bge-small-en-v1.5":{"mode":"embedding","base_model":"@cf/baai/bge-small-en-v1.5","provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"workers-ai/workers-ai/@cf/pfnet/plamo-embedding-1b":{"mode":"embedding","base_model":"@cf/pfnet/plamo-embedding-1b","provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"workers-ai/workers-ai/@cf/qwen/qwen3-embedding-0.6b":{"mode":"embedding","base_model":"@cf/qwen/qwen3-embedding-0.6b","max_input_tokens":8192,"max_tokens":8192,"provider":"workers-ai","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/"},"openai.gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock_mantle","reasoning_effort_levels":["low","medium","high","xhigh","max"],"source":"https://developers.openai.com/api/docs/models/gpt-6-astra","supports_service_tier":true,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"service_tiers":["auto","default","ultrafast"],"supports_assistant_prefill":true,"supports_reasoning_with_tool_calls":false,"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"bedrock_mantle_base_path":"openai/v1","supports_audio_input":false,"supports_audio_output":false,"supports_prompt_cache_breakpoints":true},"us.openai.gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://developers.openai.com/api/docs/models/gpt-6-astra","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-west-2","us-east-1","us-east-2","us-west-1","ca-central-1","eu-central-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":128000,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default","ultrafast"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_assistant_prefill":true},"moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"min_output_tokens":16,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_bedrock_converse_cache_point":false,"supports_prompt_cache_breakpoint":false,"supports_regex_lookaround":false,"supports_system_messages":true},"us.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","min_output_tokens":16,"max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_video_input":false,"supports_native_streaming":true},"apac.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"min_output_tokens":16,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"global.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"min_output_tokens":16,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}],"supports_video_input":false,"supports_native_streaming":true},"eu.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"min_output_tokens":16,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logit_bias","label":"Logit Bias","helpText":"Modify the likelihood of specified tokens appearing in the completion.","type":"number","range":{"min":-100,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]}]},"glm-5p3-flash":{"mode":"chat","base_model":"glm-5p3-flash","supports_reasoning":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true},"routers/glm-5p3-fast":{"mode":"chat","base_model":"routers/glm-5p3-fast","supports_reasoning":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_prompt_caching":true},"fireworks_ai/accounts/fireworks/models/glm-5p3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":128000,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":128000,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/glm-5p3-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/glm-5.3-fast","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":1048576,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p3-fast":{"mode":"chat","base_model":"glm-5.3-fast","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":1048576,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/glm-5p3-flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":1048576,"supports_reasoning":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"fast":2},"supports_output_config":true,"supports_speed":true,"supports_fast_mode":true,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/opus-5-5/overview","provider":"anthropic","web_search_billing_unit":"per_request","inference_geo_us_multiplier":1.1,"default_reasoning_effort":"medium","supports_image_input":true,"supports_parallel_function_calling":true,"supports_forced_tool_choice":false,"supports_system_messages":true,"tool_use_system_prompt_tokens":286,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_mid_conversation_system_messages":true,"supports_reasoning_disable":false,"supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"writer.palmyra-vision-7b":{"mode":"chat","base_model":"palmyra-vision-7b","max_input_tokens":4096,"max_output_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","supports_vision":true,"provider":"bedrock","max_tokens":4096,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"global.twelvelabs.pegasus-1-2-v1:0":{"mode":"chat","base_model":"pegasus-1-2","deprecation_date":"2027-03-22","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.md","supports_video_input":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"anthropic.claude-opus-5-5":{"mode":"chat","base_model":"anthropic.claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","thinking_always_on":true,"supports_forced_tool_use":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"default_reasoning_effort":"medium","supports_image_input":true,"supports_parallel_function_calling":true,"supports_forced_tool_choice":false,"supports_system_messages":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","thinking_always_on":true,"supports_forced_tool_use":false,"token_count_model":{"model":"global.anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","thinking_always_on":true,"supports_forced_tool_use":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"eu.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"au.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"jp.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":2048,"provider":"azure","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","deprecation_date":"2027-12-05","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":512,"provider":"azure","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","deprecation_date":"2027-07-08","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":512,"provider":"azure","supports_safeguards":true,"supports_mid_conversation_system_messages":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}],"server_tools":{"web_search":"web_search_20250305"}},"azure/claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"deprecation_date":"2027-09-18","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":512,"source":"https://management.azure.com/subscriptions/c873328e-b572-4770-8dff-aaeb6f1f0e79/providers/Microsoft.CognitiveServices/locations/eastus2/models?api-version=2024-10-01","provider":"azure","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}],"server_tools":{"web_search":"web_search_20250305"}},"azure/claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","deprecation_date":"2027-09-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":1024,"provider":"azure","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}],"server_tools":{"web_search":"web_search_20250305"}},"azure/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","deprecation_date":"2027-06-30","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"prompt_cache_min_tokens":1024,"provider":"azure","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}],"server_tools":{"web_search":"web_search_20250305"}},"azure/claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-chat-latest":{"mode":"chat","base_model":"gpt-chat","deprecation_date":"2026-12-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_effort_levels":["medium"],"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.5-2026-04-23":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.5-2026-04-24":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.6","max_input_tokens":922000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_computer_use":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4-2026-03-05":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4-pro":{"mode":"responses","base_model":"gpt-5.4-pro","deprecation_date":"2027-09-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"azure/gpt-5.4-pro-2026-03-05":{"mode":"responses","base_model":"gpt-5.4-pro","deprecation_date":"2027-09-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"azure/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":false,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4-mini-2026-03-17":{"mode":"chat","base_model":"gpt-5.4-mini","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":false,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":false,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.4-nano-2026-03-17":{"mode":"chat","base_model":"gpt-5.4-nano","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":false,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/model-router":{"mode":"chat","base_model":"model-router","comment":"Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/<deployment-name> where deployment-name is your Azure deployment (e.g., azure-model-router)","deprecation_date":"2027-05-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":200000,"max_output_tokens":32768,"source":"https://azure.microsoft.com/en-us/pricing/details/ai-services/","supports_vision":true,"provider":"azure","max_tokens":32768,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-audio-1.5-2026-02-23":{"mode":"chat","base_model":"gpt-audio-1.5","deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":16384,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-audio-mini":{"mode":"chat","base_model":"gpt-audio-mini","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","max_tokens":16384,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.3-codex":{"mode":"responses","base_model":"gpt-5.3-codex","deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"azure/us/gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.4-2026-03-05":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.4-2026-03-05":{"mode":"chat","base_model":"gpt-5.4","deprecation_date":"2027-09-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"provider":"azure","supports_service_tier":true,"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6":{"mode":"chat","base_model":"gpt-5.6","max_input_tokens":922000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6-sol","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-sol-2026-07-09":{"mode":"chat","base_model":"gpt-5.6-sol","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6-sol","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6.1-terra","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-terra-2026-07-09":{"mode":"chat","base_model":"gpt-5.6-terra","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6.1-terra","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6-luna","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.6-luna-2026-07-09":{"mode":"chat","base_model":"gpt-5.6-luna","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-6-luna","max_input_tokens":922000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/chat-latest":{"mode":"chat","base_model":"chat","deprecation_date":"2026-12-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_effort_levels":["medium"],"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.6":{"mode":"chat","base_model":"gpt-5.6","max_input_tokens":922000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-6-sol","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-6.1-terra","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-6-luna","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-chat-latest":{"mode":"chat","base_model":"gpt-chat","deprecation_date":"2026-12-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":272000,"max_output_tokens":128000,"reasoning_effort_levels":["medium"],"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.6":{"mode":"chat","base_model":"gpt-5.6","max_input_tokens":922000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-6-sol","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-6.1-terra","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","deprecation_date":"2028-01-11","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-6-luna","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.5-2026-04-23":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.5-2026-04-24":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.5-2026-04-23":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.5-2026-04-24":{"mode":"chat","base_model":"gpt-5.5","deprecation_date":"2027-10-26","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/gpt-5.5-pro":{"mode":"responses","base_model":"gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":false,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"azure/gpt-5.5-pro-2026-04-23":{"mode":"responses","base_model":"gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"azure/Codestral-2501":{"mode":"chat","base_model":"codestral","deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":256000,"max_output_tokens":4096,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_native_streaming":true,"supports_pdf_input":false,"provider":"azure","max_tokens":4096,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-DeepSeek-V3.2":{"mode":"chat","base_model":"fw-deepseek-v3.2","deprecation_date":"2027-07-01","max_input_tokens":163840,"max_output_tokens":163840,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":163840,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-DeepSeek-V4-Pro":{"mode":"chat","base_model":"fw-deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":384000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":384000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-GLM-5":{"mode":"chat","base_model":"fw-glm-5","deprecation_date":"2027-07-01","successor_model":"azure_ai/FW-GLM-5.2","max_input_tokens":200000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-GLM-5.1":{"mode":"chat","base_model":"fw-glm-5.1","deprecation_date":"2027-07-01","successor_model":"azure_ai/FW-GLM-5.2","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-GLM-5.2":{"mode":"chat","base_model":"fw-glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-GLM-5.2-Fast":{"mode":"chat","base_model":"fw-glm-5.2-fast","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Inkling":{"mode":"chat","base_model":"fw-inkling","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":1048576,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Kimi-K2.5":{"mode":"chat","base_model":"fw-kimi-k2.5","deprecation_date":"2027-07-01","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":262144,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Kimi-K2.6":{"mode":"chat","base_model":"fw-kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":262144,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Kimi-K2.7-Code":{"mode":"chat","base_model":"fw-kimi-k2.7-code","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":262144,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Kimi-K3":{"mode":"chat","base_model":"fw-kimi-k3","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-MiniMax-M2.5":{"mode":"chat","base_model":"fw-minimax-m2.5","deprecation_date":"2027-07-01","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":1000000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-MiniMax-M3":{"mode":"chat","base_model":"fw-minimax-m3","max_input_tokens":512000,"max_output_tokens":512000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":512000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Nemotron-Lightning-3.5-30B-A3B":{"mode":"chat","base_model":"fw-nemotron-lightning-3.5-30b-a3b","max_input_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/FW-Nemotron-3-Ultra-NVFP4":{"mode":"chat","base_model":"fw-nemotron-3-ultra-nvfp4","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supported_endpoints":["/v1/chat/completions"],"max_tokens":262144,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"azure/MAI-Thinking-1":{"mode":"chat","base_model":"mai-thinking-1","max_input_tokens":256000,"max_output_tokens":64000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","max_tokens":64000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/cohere-command-a":{"mode":"chat","base_model":"command-a","deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"coherelabs-command-a-plus-05-2026-fp81","max_input_tokens":131072,"max_output_tokens":8182,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"provider":"azure","max_tokens":8182,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8182},"supported_endpoints":["/v1/chat/completions"]}]},"azure/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","deprecation_date":"2028-02-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":1000000,"max_output_tokens":384000,"reasoning_effort_levels":["none","high","max"],"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_prompt_caching":true,"provider":"azure","max_tokens":384000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2028-02-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/DeepSeek-V4-Flash-0731","max_input_tokens":1000000,"max_output_tokens":384000,"reasoning_effort_levels":["none","high","max"],"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_prompt_caching":true,"provider":"azure","max_tokens":384000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","deprecation_date":"2026-12-03","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":1000000,"max_output_tokens":384000,"reasoning_effort_levels":["none","high","max"],"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/deepseek/","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supports_web_search":true,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_assistant_prefill":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_system_messages":true,"max_tokens":384000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4.3":{"mode":"chat","base_model":"grok-4.3","max_input_tokens":20000,"max_output_tokens":8192,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":8192,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":200000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4-20-reasoning":{"mode":"chat","base_model":"grok-4-20","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":262000,"max_output_tokens":8192,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_reasoning":true,"provider":"azure","max_tokens":8192,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4-20-non-reasoning":{"mode":"chat","base_model":"grok-4-20-non","deprecation_date":"2027-04-06","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":262000,"max_output_tokens":8192,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","max_tokens":8192,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4-1-fast-non-reasoning":{"mode":"chat","base_model":"grok-4-1-fast-non","deprecation_date":"2027-05-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/grok-4-1-fast-reasoning":{"mode":"chat","base_model":"grok-4-1-fast","deprecation_date":"2027-05-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","token_count_framing":149,"token_count_tool_framing":302,"token_count_responses_framing":145,"token_count_responses_tool_framing":100,"max_input_tokens":131072,"max_output_tokens":131072,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"azure","max_tokens":131072,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]}]},"azure/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","deprecation_date":"2027-04-16","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-kimi-k2-6-in-microsoft-foundry/4513125","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_prompt_caching":true,"provider":"azure","max_tokens":262144,"supports_video_input":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"bedrock/ap-northeast-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/ap-south-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/ap-southeast-2/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/ap-southeast-3/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-north-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-central-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-west-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-west-2/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-west-2/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nvidia-nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-south-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/sa-east-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-east-1/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-east-2/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","max_input_tokens":300000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/amazon.nova-micro-v1:0":{"mode":"chat","base_model":"nova-micro","max_input_tokens":128000,"max_output_tokens":10000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":10000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-west-2/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":1000000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cerebras/qwen-3.8-27b":{"mode":"chat","base_model":"qwen3.8-27b","max_input_tokens":65536,"max_output_tokens":32768,"source":"https://api.cerebras.ai/public/v1/models/qwen-3.8-27b","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"cerebras","supports_developer_role":false,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-opus-4-6-20260205":{"mode":"chat","base_model":"claude-opus-4-6","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"provider_specific_entry":{"us":1.1,"fast":6},"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_speed":true,"prompt_cache_min_tokens":4096,"provider":"anthropic","web_search_billing_unit":"per_request","supports_web_search":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-opus-4-7-20260416":{"mode":"chat","base_model":"claude-opus-4-7","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1,"fast":6},"supports_output_config":true,"prompt_cache_min_tokens":2048,"provider":"anthropic","web_search_billing_unit":"per_request","deprecation_date":"2027-04-16","supports_web_search":true,"supports_safeguards":true,"max_tokens":128000,"supports_speed":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/openai/gpt-oss-120b":{"mode":"chat","base_model":"","max_input_tokens":128000,"max_output_tokens":128000,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/google/gemma-2b-it-lora":{"mode":"chat","base_model":"","max_input_tokens":8192,"max_output_tokens":8192,"rpm":300,"provider":"cloudflare","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-3.2-3b-instruct":{"mode":"chat","base_model":"","max_input_tokens":80000,"max_output_tokens":80000,"rpm":300,"provider":"cloudflare","max_tokens":80000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-guard-3-8b":{"mode":"chat","base_model":"","max_input_tokens":131072,"max_output_tokens":131072,"rpm":300,"provider":"cloudflare","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/mistral/mistral-7b-instruct-v0.2-lora":{"mode":"chat","base_model":"","max_input_tokens":15000,"max_output_tokens":15000,"rpm":300,"provider":"cloudflare","max_tokens":15000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":15000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/moonshotai/kimi-k2.7-code":{"mode":"chat","base_model":"","max_input_tokens":262144,"max_output_tokens":262144,"rpm":20,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/deepseek-ai/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"","max_input_tokens":80000,"max_output_tokens":80000,"rpm":300,"supports_reasoning":true,"provider":"cloudflare","max_tokens":80000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-3.1-8b-instruct-fp8":{"mode":"chat","base_model":"","max_input_tokens":32000,"max_output_tokens":32000,"rpm":300,"provider":"cloudflare","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-3.2-1b-instruct":{"mode":"chat","base_model":"","max_input_tokens":60000,"max_output_tokens":60000,"rpm":300,"provider":"cloudflare","max_tokens":60000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":60000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/moonshotai/kimi-k2.6":{"mode":"chat","base_model":"","max_input_tokens":262144,"max_output_tokens":262144,"rpm":20,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/zai-org/glm-4.7-flash":{"mode":"chat","base_model":"","max_input_tokens":131072,"max_output_tokens":131072,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta-llama/llama-2-7b-chat-hf-lora":{"mode":"chat","base_model":"","max_input_tokens":8192,"max_output_tokens":8192,"rpm":300,"provider":"cloudflare","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-3.3-70b-instruct-fp8-fast":{"mode":"chat","base_model":"","max_input_tokens":24000,"max_output_tokens":24000,"rpm":300,"supports_function_calling":true,"provider":"cloudflare","max_tokens":24000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/ibm-granite/granite-4.0-h-micro":{"mode":"chat","base_model":"","max_input_tokens":131000,"max_output_tokens":131000,"rpm":300,"supports_function_calling":true,"provider":"cloudflare","max_tokens":131000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/qwen/qwen2.5-coder-32b-instruct":{"mode":"chat","base_model":"","max_input_tokens":32768,"max_output_tokens":32768,"rpm":300,"provider":"cloudflare","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/zai-org/glm-5.2":{"mode":"chat","base_model":"","max_input_tokens":262144,"max_output_tokens":262144,"rpm":20,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/nvidia/nemotron-3-120b-a12b":{"mode":"chat","base_model":"","max_input_tokens":256000,"max_output_tokens":256000,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"mode":"chat","base_model":"","max_input_tokens":128000,"max_output_tokens":128000,"rpm":300,"provider":"cloudflare","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/qwen/qwen3-30b-a3b-fp8":{"mode":"chat","base_model":"","max_input_tokens":32768,"max_output_tokens":32768,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/google/gemma-7b-it-lora":{"mode":"chat","base_model":"","max_input_tokens":3500,"max_output_tokens":3500,"rpm":300,"provider":"cloudflare","max_tokens":3500,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3500,"range":{"min":1,"max":3500},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"","max_input_tokens":256000,"max_output_tokens":256000,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/mistralai/mistral-small-3.1-24b-instruct":{"mode":"chat","base_model":"","max_input_tokens":128000,"max_output_tokens":128000,"rpm":300,"supports_function_calling":true,"provider":"cloudflare","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-3.2-11b-vision-instruct":{"mode":"chat","base_model":"","max_input_tokens":128000,"max_output_tokens":128000,"rpm":300,"supports_vision":true,"provider":"cloudflare","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/openai/gpt-oss-20b":{"mode":"chat","base_model":"","max_input_tokens":128000,"max_output_tokens":128000,"rpm":300,"supports_function_calling":true,"supports_reasoning":true,"provider":"cloudflare","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/meta/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"","max_input_tokens":131000,"max_output_tokens":131000,"rpm":300,"supports_function_calling":true,"provider":"cloudflare","max_tokens":131000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/@cf/qwen/qwq-32b":{"mode":"chat","base_model":"","max_input_tokens":24000,"max_output_tokens":24000,"rpm":300,"supports_reasoning":true,"provider":"cloudflare","max_tokens":24000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"command-a-plus-05-2026":{"mode":"chat","base_model":"command-a-plus","max_input_tokens":128000,"max_output_tokens":64000,"source":"https://docs.cohere.com/docs/command-a-plus","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"cohere","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/glm-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202745,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/glm-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":229376,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-max-2026-01-23":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-vl-32b-instruct":{"mode":"chat","base_model":"qwen3-vl-32b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-vl-32b-thinking":{"mode":"chat","base_model":"qwen3-vl-32b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3-vl-plus":{"mode":"chat","base_model":"qwen3-vl-plus","max_input_tokens":260096,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":2e-7,"output_cost_per_token":0.0000016,"range":[0,32000]},{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000024,"range":[32000,128000]},{"input_cost_per_token":6e-7,"output_cost_per_token":0.0000048,"range":[128000,256000]}],"provider":"dashscope","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.5-plus":{"mode":"chat","base_model":"qwen3.5-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_token":0.0000024,"range":[0,256000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.000003,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":4e-7,"output_cost_per_token":0.0000016,"range":[0,256000]},{"cache_read_input_token_cost":2.4e-7,"input_cost_per_token":0.0000012,"output_cost_per_token":0.0000048,"range":[256000,1000000]}],"provider":"dashscope","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"dashscope","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.8-flash":{"mode":"chat","base_model":"qwen3.8-flash","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://docs.modelstudio.console.alibabacloud.com/en/model-studio/model-pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"dashscope","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"dashscope/qwen3.8-omni-flash":{"mode":"chat","base_model":"qwen3.8-omni-flash","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://docs.modelstudio.console.alibabacloud.com/en/model-studio/model-pricing","supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"dashscope","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/glm-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202745,"max_output_tokens":131072,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/glm-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":229376,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-coder":{"mode":"chat","base_model":"qwen-coder","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-flash":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-flash-2025-07-28":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-max":{"mode":"chat","base_model":"qwen-max","max_input_tokens":30720,"max_output_tokens":8192,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-2025-01-25":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":8192,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-2025-04-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-2025-07-14":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-2025-07-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-2025-09-11":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-plus-latest":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-turbo":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-turbo-2024-11-01":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-turbo-2025-04-28":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen-turbo-latest":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-30b-a3b":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-coder-flash":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"cache_read_input_token_cost":1.2e-7,"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"cache_read_input_token_cost":2e-7,"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"cache_read_input_token_cost":4e-7,"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-coder-flash-2025-07-28":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-coder-plus":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":1e-7,"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"cache_read_input_token_cost":1.8e-7,"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"cache_read_input_token_cost":3e-7,"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"cache_read_input_token_cost":6e-7,"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-coder-plus-2025-07-22":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-max-preview":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-max":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-max-2026-01-23":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-vl-32b-instruct":{"mode":"chat","base_model":"qwen3-vl-32b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-vl-32b-thinking":{"mode":"chat","base_model":"qwen3-vl-32b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3-vl-plus":{"mode":"chat","base_model":"qwen3-vl-plus","max_input_tokens":260096,"max_output_tokens":32768,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":2e-7,"output_cost_per_token":0.0000016,"range":[0,32000]},{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000024,"range":[32000,128000]},{"input_cost_per_token":6e-7,"output_cost_per_token":0.0000048,"range":[128000,256000]}],"provider":"qwencloud","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3.5-plus":{"mode":"chat","base_model":"qwen3.5-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_token":0.0000024,"range":[0,256000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.000003,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":4e-7,"output_cost_per_token":0.0000016,"range":[0,256000]},{"cache_read_input_token_cost":2.4e-7,"input_cost_per_token":0.0000012,"output_cost_per_token":0.0000048,"range":[256000,1000000]}],"provider":"qwencloud","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwencloud","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwencloud/qwq-plus":{"mode":"chat","base_model":"qwq-plus","max_input_tokens":98304,"max_output_tokens":8192,"source":"https://www.qwencloud.com/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwencloud","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":393216,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/glm-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202745,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/glm-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":229376,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-coder":{"mode":"chat","base_model":"qwen-coder","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-flash":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-flash-2025-07-28":{"mode":"chat","base_model":"qwen-flash","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":5e-8,"output_cost_per_token":4e-7,"range":[0,256000]},{"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-max":{"mode":"chat","base_model":"qwen-max","max_input_tokens":30720,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-2025-01-25":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-2025-04-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-2025-07-14":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-2025-07-28":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-2025-09-11":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-plus-latest":{"mode":"chat","base_model":"qwen-plus","max_input_tokens":997952,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_reasoning_token":0.000004,"output_cost_per_token":0.0000012,"range":[0,256000]},{"input_cost_per_token":0.0000012,"output_cost_per_reasoning_token":0.000012,"output_cost_per_token":0.0000036,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-turbo":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-turbo-2024-11-01":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-turbo-2025-04-28":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen-turbo-latest":{"mode":"chat","base_model":"qwen-turbo","max_input_tokens":1000000,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-30b-a3b":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":129024,"max_output_tokens":16384,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-coder-flash":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"cache_read_input_token_cost":1.2e-7,"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"cache_read_input_token_cost":2e-7,"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"cache_read_input_token_cost":4e-7,"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-coder-flash-2025-07-28":{"mode":"chat","base_model":"qwen3-coder-flash","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000015,"range":[0,32000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"range":[32000,128000]},{"input_cost_per_token":8e-7,"output_cost_per_token":0.000004,"range":[128000,256000]},{"input_cost_per_token":0.0000016,"output_cost_per_token":0.0000096,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-coder-plus":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"cache_read_input_token_cost":1e-7,"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"cache_read_input_token_cost":1.8e-7,"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"cache_read_input_token_cost":3e-7,"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"cache_read_input_token_cost":6e-7,"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-coder-plus-2025-07-22":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":997952,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.000001,"output_cost_per_token":0.000005,"range":[0,32000]},{"input_cost_per_token":0.0000018,"output_cost_per_token":0.000009,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,256000]},{"input_cost_per_token":0.000006,"output_cost_per_token":0.00006,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-max-preview":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-max":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-max-2026-01-23":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":258048,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"tiered_pricing":[{"input_cost_per_token":0.0000012,"output_cost_per_token":0.000006,"range":[0,32000]},{"input_cost_per_token":0.0000024,"output_cost_per_token":0.000012,"range":[32000,128000]},{"input_cost_per_token":0.000003,"output_cost_per_token":0.000015,"range":[128000,252000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-vl-32b-instruct":{"mode":"chat","base_model":"qwen3-vl-32b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-vl-32b-thinking":{"mode":"chat","base_model":"qwen3-vl-32b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3-vl-plus":{"mode":"chat","base_model":"qwen3-vl-plus","max_input_tokens":260096,"max_output_tokens":32768,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":2e-7,"output_cost_per_token":0.0000016,"range":[0,32000]},{"input_cost_per_token":3e-7,"output_cost_per_token":0.0000024,"range":[32000,128000]},{"input_cost_per_token":6e-7,"output_cost_per_token":0.0000048,"range":[128000,256000]}],"provider":"qwen_ai_platform","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.5-plus":{"mode":"chat","base_model":"qwen3.5-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":4e-7,"output_cost_per_token":0.0000024,"range":[0,256000]},{"input_cost_per_token":5e-7,"output_cost_per_token":0.000003,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":991808,"max_output_tokens":65536,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tiered_pricing":[{"cache_read_input_token_cost":8e-8,"input_cost_per_token":4e-7,"output_cost_per_token":0.0000016,"range":[0,256000]},{"cache_read_input_token_cost":2.4e-7,"input_cost_per_token":0.0000012,"output_cost_per_token":0.0000048,"range":[256000,1000000]}],"provider":"qwen_ai_platform","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"qwen_ai_platform","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.8-flash":{"mode":"chat","base_model":"qwen3.8-flash","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://docs.modelstudio.console.alibabacloud.com/en/model-studio/model-pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"qwen_ai_platform","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwen3.8-omni-flash":{"mode":"chat","base_model":"qwen3.8-omni-flash","max_input_tokens":991808,"max_output_tokens":131072,"source":"https://docs.modelstudio.console.alibabacloud.com/en/model-studio/model-pricing","supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"qwen_ai_platform","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"qwen_ai_platform/qwq-plus":{"mode":"chat","base_model":"qwq-plus","max_input_tokens":98304,"max_output_tokens":8192,"source":"https://www.alibabacloud.com/help/en/model-studio/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"qwen_ai_platform","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Text-only input. Prompts and responses retained 30 days for trust and safety. In-geo endpoint is 10% higher."},"prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":false,"thinking_always_on":true,"provider":"databricks","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"databricks","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_legacy_thinking":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":4096,"provider":"databricks","max_tokens":64000,"supports_minimal_reasoning_effort":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"prompt_cache_min_tokens":2048,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","max_tokens":64000,"supports_minimal_reasoning_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"prompt_cache_min_tokens":1024,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","max_tokens":1000000,"supports_minimal_reasoning_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Billed under the 'Claude Opus 4.5 / 4.6 / 4.7 / 4.8 / 5' row. In-geo endpoint is 10% higher."},"prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_legacy_thinking":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":1024,"provider":"databricks","max_tokens":64000,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","max_input_tokens":200000,"max_output_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. INTRODUCTORY PRICING through Aug 31, 2026. After that, Sonnet 5 reverts to the Sonnet 4.5/4.6 rates (42.857 input / 214.286 output DBU per 1M). Does not support temperature, top_p or top_k."},"prompt_cache_min_tokens":1024,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":true,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-11-05","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","token_count_framing":97,"max_input_tokens":1000000,"max_output_tokens":393216,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Billing reads the per-token dollar fields; the '*_dbu_cost_per_token' fields are the published Databricks rates, kept for reference. Context/max output are the DeepSeek-published model limits (1M context, 384K max output)."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","successor_model":"databricks/databricks-deepseek-v4-1-flash-1017","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"databricks","is_deprecated":true,"max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","deprecation_date":"2026-10-30","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","max_input_tokens":1000000,"max_output_tokens":393216,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Billing reads the per-token dollar fields; the '*_dbu_cost_per_token' fields are the published Databricks rates, kept for reference. Context/max output are the DeepSeek-published model limits (1M context, 384K max output)."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","successor_model":"databricks/databricks-deepseek-v4-1-flash-1017","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"databricks","is_deprecated":true,"max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Rates exclude a 20% promotional discount in effect until Jan 31, 2027. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-1-flash-image":{"mode":"chat","base_model":"gemini-3.1-flash-image","max_input_tokens":131072,"max_output_tokens":32768,"metadata":{"notes":"Databricks DBU rates not yet published for this model; endpoint metadata only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_vision":true,"provider":"databricks","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-pro-image":{"mode":"chat","base_model":"gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"metadata":{"notes":"Databricks DBU rates not yet published for this model; endpoint metadata only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_vision":true,"provider":"databricks","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-1-pro":{"mode":"chat","base_model":"gemini-3.1-pro","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Billed under the 'Gemini 3.0 / 3.1 Pro' row. Long context (>200k tokens) tier: 71.429 input / 321.429 output DBU per 1M. Rates exclude a 20% promotional discount until Jan 31, 2027."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-flash":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Billed under the 'Gemini 3.0 Flash' row. Rates exclude a 20% promotional discount until Jan 31, 2027."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-pro":{"mode":"chat","base_model":"gemini-3-pro","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Databricks DBU rates not yet published for this model; endpoint metadata only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Databricks DBU rates not yet published for this model; endpoint metadata only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Rates exclude a 20% promotional discount in effect until Jan 31, 2027 (actual billed price is 20% lower)."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"deprecation_date":"2026-11-19","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/retired-models-policy","successor_model":"databricks/databricks-gemini-3-8-flash","supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Rates exclude a 20% promotional discount in effect until Jan 31, 2027. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gemini-3-5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Rates exclude a 20% promotional discount in effect until Jan 31, 2027. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":true,"supports_video_input":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-glm-5-2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1000000,"max_output_tokens":131072,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Pay-per-token only; no provisioned throughput. 1M token context per Databricks docs."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"deprecation_date":"2027-02-04","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","successor_model":"databricks/databricks-glm-5-3","provider":"databricks","supports_assistant_prefill":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-glm-5-3":{"mode":"chat","base_model":"glm-5-3","max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"thinking_always_on":true,"provider":"databricks","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-glm-5-3-flash":{"mode":"chat","base_model":"glm-5-3-flash","max_input_tokens":1048576,"max_output_tokens":131072,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"databricks","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-2":{"mode":"chat","base_model":"gpt-5.2","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Billed under the 'GPT 5.2/5.3 Codex' row. Responses API only; no batch inference."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"provider":"databricks","supports_assistant_prefill":true,"supports_reasoning":true,"supports_vision":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-4":{"mode":"chat","base_model":"gpt-5.4","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Long context (>200k input) tier: 71.428 input / 321.429 output DBU per 1M."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Long context (>200k input) tier: 142.857 input / 642.857 output DBU per 1M. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Long context (>200k input) tier: 57.143 input / 257.143 output DBU per 1M. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Long context (>200k input) tier: 5.714 input / 25.714 output DBU per 1M. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-5":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Long context (>200k input) tier: 142.857 input / 642.857 output DBU per 1M. Responses API only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-gpt-5-5-pro":{"mode":"chat","base_model":"gpt-5.5-pro","max_input_tokens":400000,"max_output_tokens":128000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Billed under the 'GPT 5.4/5.5 Pro' row. No cache-read tier. Long context tier: 857.142 input / 3857.144 output DBU per 1M. Responses API only."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-grok-4-6":{"mode":"chat","base_model":"grok-4-6","max_input_tokens":500000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"databricks","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-inkling":{"mode":"chat","base_model":"inkling","deprecation_date":"2026-10-30","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","max_input_tokens":1000000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Public Preview. Pay-per-token only. 1M token context per Databricks docs."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","successor_model":"databricks/databricks-glm-5-3","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"databricks","supports_assistant_prefill":true,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-kimi-k3":{"mode":"chat","base_model":"kimi-k3","token_count_framing":100,"max_input_tokens":1000000,"max_output_tokens":1048576,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Pay-per-token only; no provisioned throughput. 1M token context per Databricks docs."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","supports_assistant_prefill":true,"max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-qwen35-122b-a10b":{"mode":"chat","base_model":"qwen3.5-122b-a10b","max_input_tokens":256000,"max_output_tokens":25000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Public Preview. Reasoning-only model; reasoning cannot be disabled. A 'Priority' tier is billed at 2x (6.286 input / 62.858 output DBU per 1M)."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"thinking_always_on":true,"provider":"databricks","supports_assistant_prefill":true,"supports_prompt_caching":false,"max_tokens":25000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":25000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. Public Preview."},"source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":false,"provider":"databricks","max_input_tokens":262144,"supports_assistant_prefill":true,"supports_prompt_caching":false,"supports_reasoning":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/nvidia/NVIDIA-Nemotron-3.5-Lightning":{"mode":"chat","base_model":"nvidia-nemotron-3.5-lightning","max_input_tokens":262144,"source":"https://deepinfra.com/pricing","supports_tool_choice":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"provider":"deepinfra","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fal_ai/fal-ai/moondream3-preview/query":{"mode":"chat","base_model":"moondream3-preview/query","source":"https://fal.ai/models/fal-ai/moondream3-preview/query","supported_endpoints":["/v1/chat/completions"],"supports_reasoning":true,"supports_vision":true,"provider":"fal_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-08-14","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":1048576,"max_output_tokens":384000,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro-0813","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/glm-4p7":{"mode":"chat","base_model":"glm-4.7","max_input_tokens":202800,"max_output_tokens":202800,"source":"https://fireworks.ai/models/fireworks/glm-4p7","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":202800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/glm-5p1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/glm-5p2":{"mode":"chat","base_model":"glm-5.2","deprecation_date":"2026-09-25","successor_model":"fireworks_ai/accounts/fireworks/models/glm-5p3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2p6":{"mode":"chat","base_model":"kimi-k2.6","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k2p7-code":{"mode":"chat","base_model":"kimi-k2.7-code","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/minimax-m2p1":{"mode":"chat","base_model":"minimax-m2.1","max_input_tokens":204800,"max_output_tokens":204800,"source":"https://fireworks.ai/models/fireworks/minimax-m2p1","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":204800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/minimax-m3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":512000,"max_output_tokens":512000,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":512000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-08-14","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":1048576,"max_output_tokens":384000,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/deepseek-v4p1-flash","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-4p7":{"mode":"chat","base_model":"glm-4.7","max_input_tokens":202800,"max_output_tokens":202800,"source":"https://fireworks.ai/models/fireworks/glm-4p7","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":202800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p1-fast":{"mode":"chat","base_model":"glm-5.1-fast","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p2":{"mode":"chat","base_model":"glm-5.2","deprecation_date":"2026-09-25","successor_model":"fireworks_ai/glm-5p3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":32768,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k2p5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://fireworks.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k2p6":{"mode":"chat","base_model":"kimi-k2.6","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k2p6-fast":{"mode":"chat","base_model":"kimi-k2.6-fast","deprecation_date":"2026-08-27","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/kimi-k2p6","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k2p7-code":{"mode":"chat","base_model":"kimi-k2.7-code","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k2p7-code-fast":{"mode":"chat","base_model":"kimi-k2.7-code-fast","deprecation_date":"2026-08-27","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/kimi-k2p7-code","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/minimax-m2p1":{"mode":"chat","base_model":"minimax-m2.1","max_input_tokens":204800,"max_output_tokens":204800,"source":"https://fireworks.ai/models/fireworks/minimax-m2p1","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","max_tokens":204800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/minimax-m3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":512000,"max_output_tokens":512000,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":512000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/qwen3p7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_image_input":true,"supports_video_input":true,"comment":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/zai-org/GLM-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"comment":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":262144,"max_output_tokens":262144,"supports_prompt_caching":false,"supports_reasoning":true,"reasoning_effort_levels":[],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_image_input":true,"supports_video_input":false,"comment":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":["high","max"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"comment":"Open flagship GLM for long-horizon coding agents and million-token context work","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/deepseek-ai/DeepSeek-V3.2":{"mode":"chat","base_model":"deepseek-v3.2","max_input_tokens":163840,"max_output_tokens":163840,"max_tokens":163840,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":[],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"comment":"DeepSeek chat model for instruction following, coding, and analysis","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/MiniMaxAI/MiniMax-M2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":196608,"max_output_tokens":196608,"max_tokens":196608,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":[],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"comment":"Prior MiniMax coding model for agent workflows, office edits, and automation","source":"https://api.friendli.ai/serverless/v1/models","provider":"friendliai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"friendliai/zai-org/GLM-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202752,"max_output_tokens":202752,"supports_prompt_caching":true,"supports_reasoning":true,"reasoning_effort_levels":[],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"comment":"Strong GLM coding model for agentic engineering, terminals, and repository generation","source":"https://api.friendli.ai/serverless/v1/models","deprecation_date":"2026-11-20","provider":"friendliai","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","max_input_tokens":1048576,"max_output_tokens":65536,"prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-flash","successor_model":"vertex_ai/gemini-3.8-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":65536,"supports_audio_output":false,"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"service_tiers":["priority","flex"],"deprecation_date":"2027-05-19","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"regional_endpoint_uplift_multiplier":1.1,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.6-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"deprecation_date":"2026-11-19","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/gemini-3.8-flash","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","supports_audio_output":false,"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"service_tiers":["priority","flex"],"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"regional_endpoint_uplift_multiplier":1.1,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.7-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"deprecation_date":"2027-01-28","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/gemini-3.8-flash","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":65536,"supports_audio_output":false,"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"service_tiers":["priority","flex"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","max_input_tokens":1048576,"max_output_tokens":65536,"prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","service_tiers":["priority","flex"],"supports_code_execution":true,"supports_service_tier":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.1-pro-preview-customtools":{"mode":"chat","base_model":"gemini-3.1-pro-customtools","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_url_context":true,"supports_native_streaming":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-robotics-er-2-preview":{"mode":"chat","base_model":"gemini-robotics-er-2","max_input_tokens":131072,"max_output_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3-pro-image":{"mode":"image_generation","base_model":"gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":false,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","supports_reasoning":false,"provider":"gemini","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/nano-banana-pro-preview":{"mode":"image_generation","base_model":"nano-banana-pro","max_input_tokens":131072,"max_output_tokens":32768,"rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.1-flash-image":{"mode":"image_generation","base_model":"gemini-3.1-flash-image","max_input_tokens":131072,"max_output_tokens":32768,"rpm":1000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://ai.google.dev/gemini-api/docs/pricing","successor_model":"gemini/gemini-nano-banana-2.1","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.1-flash-lite-image":{"mode":"image_generation","base_model":"gemini-3.1-flash-lite-image","max_input_tokens":65536,"max_output_tokens":66000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-3.1-flash-lite-image","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":false,"provider":"gemini","max_tokens":66000,"rpm":null,"tpm":null,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":66000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-flash","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","successor_model":"gemini/gemini-3.6-flash","provider":"gemini","max_tokens":65536,"supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"gemini","supports_code_execution":true,"supports_file_search":true,"supports_service_tier":true,"supports_computer_use":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/models/gemini-3.7-flash?s","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","successor_model":"gemini/gemini-3.8-flash","provider":"gemini","max_tokens":65536,"supports_code_execution":true,"supports_computer_use":true,"supports_file_search":true,"supports_service_tier":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","supports_computer_use":true,"provider":"gemini","service_tiers":["priority","flex"],"supports_code_execution":true,"supports_service_tier":true,"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.1-pro-preview-customtools":{"mode":"chat","base_model":"gemini-3.1-pro-customtools","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_url_context":true,"supports_native_streaming":true,"tpm":800000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini-omni-flash-preview":{"mode":"chat","base_model":"gemini-omni-flash","max_input_tokens":1048576,"max_output_tokens":65535,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","video"],"supports_audio_input":true,"supports_reasoning":true,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"deprecation_date":"2026-11-06","provider":"vertex_ai","is_deprecated":true,"max_tokens":65535,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","max_input_tokens":262144,"max_output_tokens":32768,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://ai.google.dev/gemini-api/docs/pricing","provider":"gemini","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemma-4-31b-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":262144,"max_output_tokens":32768,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://ai.google.dev/gemini-api/docs/pricing","provider":"gemini","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/lyria-3-clip-preview":{"mode":"chat","base_model":"lyria-3-clip","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/lyria-3-pro-preview":{"mode":"chat","base_model":"lyria-3-pro","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"successor_model":"gemini/lyria-3.5","provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"github_copilot/claude-opus-4.6-fast":{"mode":"chat","base_model":"claude-opus-4-6-fast","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"max_input_tokens":128000,"max_output_tokens":16000,"max_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"provider":"github_copilot","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"github_copilot/gpt-5.3-codex":{"mode":"responses","base_model":"gpt-5.3-codex","max_input_tokens":128000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"github_copilot/mai-code-1-flash":{"mode":"chat","base_model":"mai-code-1-flash","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"provider":"github_copilot","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"github_copilot/mai-code-1-flash-internal":{"mode":"chat","base_model":"mai-code-1-flash-internal","max_input_tokens":128000,"max_output_tokens":64000,"max_tokens":64000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"provider":"github_copilot","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"chatgpt/gpt-5.5":{"mode":"responses","base_model":"gpt-5.5","source":"https://platform.openai.com/docs/models/gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.6-luna":{"mode":"responses","base_model":"gpt-5.6-luna","source":"https://platform.openai.com/docs/models/gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_reasoning":true,"supports_max_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.6-sol":{"mode":"responses","base_model":"gpt-5.6-sol","source":"https://platform.openai.com/docs/models/gpt-5.6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_reasoning":true,"supports_max_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.6-terra":{"mode":"responses","base_model":"gpt-5.6-terra","source":"https://platform.openai.com/docs/models/gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_reasoning":true,"supports_max_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.4":{"mode":"responses","base_model":"gpt-5.4","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"supports_none_reasoning_effort":true,"default_reasoning_effort":"none","supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.4-pro":{"mode":"responses","base_model":"gpt-5.4-pro","max_input_tokens":1050000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.3-codex":{"mode":"responses","base_model":"gpt-5.3-codex","max_input_tokens":128000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.3-codex-spark":{"mode":"responses","base_model":"gpt-5.3-codex-spark","max_input_tokens":128000,"max_output_tokens":128000,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"chatgpt","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.3-instant":{"mode":"responses","base_model":"gpt-5.3-instant","max_input_tokens":128000,"max_output_tokens":64000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"chatgpt","max_tokens":64000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/responses"]}]},"chatgpt/gpt-5.3-chat-latest":{"mode":"responses","base_model":"gpt-5.3-chat","max_input_tokens":128000,"max_output_tokens":64000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_reasoning":true,"provider":"chatgpt","max_tokens":64000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/responses"]}]},"gigachat/GigaChat-2":{"mode":"chat","base_model":"gigachat-2","max_input_tokens":128000,"max_output_tokens":8192,"supports_function_calling":true,"supports_system_messages":true,"provider":"gigachat","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/MiniMaxAI/MiniMax-M2.5":{"mode":"chat","base_model":"minimax-m2.5","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/nvidia/Nemotron-120B-A12B":{"mode":"chat","base_model":"nemotron-120b-a12b","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-5":{"mode":"chat","base_model":"glm-5","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-4.7":{"mode":"chat","base_model":"glm-4.7","deprecation_date":"2026-10-16","max_input_tokens":200000,"max_output_tokens":200000,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","is_deprecated":true,"max_tokens":200000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-4.6":{"mode":"chat","base_model":"glm-4.6","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"kimi-k2.5","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K2-Thinking":{"mode":"chat","base_model":"kimi-k2-thinking","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K2-Instruct-0905":{"mode":"chat","base_model":"kimi-k2-instruct","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128072,"max_output_tokens":128072,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","max_tokens":128072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"deepseek-v3.1","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","deprecation_date":"2026-10-16","provider":"baseten","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-audio-1.5":{"mode":"chat","base_model":"gpt-audio-1.5","max_input_tokens":128000,"max_output_tokens":16384,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":16384,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-image-2.5-flare":{"mode":"image_generation","base_model":"gpt-image-2.5-flare","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/responses"],"supports_vision":true,"supports_pdf_input":true,"source":"https://developers.openai.com/api/docs/models/gpt-image-2.5-flare","provider":"openai","web_search_billing_unit":"per_request","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"supports_prompt_caching":true,"supports_function_calling":false,"supports_tool_choice":false,"supports_response_schema":false,"supports_native_streaming":false,"supports_reasoning":false,"metadata":{"notes":"Fastest GPT Image 2.5 variant for everyday generation. Quality settings: low, medium, high, xhigh, max, auto. Image cached input is $2.00/1M (0.000002 per token); text output is not billed. Token rates match gpt-image-2. No batch pricing published for this model. Rate limits are tier-based (Tier 1: 100k TPM / 5 IPM, Tier 5: 8M TPM / 250 IPM); Free tier not supported. Released 2026-09-08."},"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-image-2.5-flare-2026-09-08":{"mode":"image_generation","base_model":"gpt-image-2.5-flare","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/responses"],"supports_vision":true,"supports_pdf_input":true,"source":"https://developers.openai.com/api/docs/models/gpt-image-2.5-flare","provider":"openai","web_search_billing_unit":"per_request","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"supports_prompt_caching":true,"supports_function_calling":false,"supports_tool_choice":false,"supports_response_schema":false,"supports_native_streaming":false,"supports_reasoning":false,"metadata":{"notes":"Dated snapshot of gpt-image-2.5-flare. Same pricing and capabilities as the undated alias. Image cached input is $2.00/1M (0.000002 per token)."},"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-image-2.5-sunburst":{"mode":"image_generation","base_model":"gpt-image-2.5-sunburst","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/responses"],"supports_vision":true,"supports_pdf_input":true,"source":"https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst","provider":"openai","web_search_billing_unit":"per_request","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"supports_prompt_caching":true,"supports_function_calling":false,"supports_tool_choice":false,"supports_response_schema":false,"supports_native_streaming":false,"supports_reasoning":false,"metadata":{"notes":"Most capable GPT Image 2.5 variant (editing precision). Quality settings: low, medium, high, xhigh, max, auto. Image cached input is $2.00/1M (0.000002 per token); text output is not billed. Token rates match gpt-image-2. No batch pricing published for this model. Rate limits are tier-based (Tier 1: 100k TPM / 5 IPM, Tier 5: 8M TPM / 250 IPM); Free tier not supported. Released 2026-09-08."},"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-image-2.5-sunburst-2026-09-08":{"mode":"image_generation","base_model":"gpt-image-2.5-sunburst","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/responses"],"supports_vision":true,"supports_pdf_input":true,"source":"https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst","provider":"openai","web_search_billing_unit":"per_request","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"supports_prompt_caching":true,"supports_function_calling":false,"supports_tool_choice":false,"supports_response_schema":false,"supports_native_streaming":false,"supports_reasoning":false,"metadata":{"notes":"Dated snapshot of gpt-image-2.5-sunburst. Same pricing and capabilities as the undated alias. Image cached input is $2.00/1M (0.000002 per token)."},"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6-sol","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_native_structured_output":true,"supports_file_search":true,"supports_service_tier":true,"metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default","flex","priority","fast"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6-luna","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"supports_decisions":true,"regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_native_structured_output":true,"supports_file_search":true,"supports_service_tier":true,"metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default","flex","priority","fast"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-5.6-cyber":{"mode":"chat","base_model":"gpt-5.6-cyber","max_input_tokens":400000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/pricing","supports_computer_use":true,"supports_parallel_function_calling":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"daybreak-red-latest":{"mode":"chat","base_model":"daybreak-red","max_input_tokens":400000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-daybreak-red-latest","supports_computer_use":true,"supports_parallel_function_calling":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-daybreak-red-latest":{"mode":"responses","base_model":"gpt-daybreak-red","max_input_tokens":400000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-daybreak-red-latest","supports_computer_use":true,"supports_parallel_function_calling":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"daybreak-blue-latest":{"mode":"chat","base_model":"daybreak-blue","max_input_tokens":922000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-daybreak-blue-latest","supports_parallel_function_calling":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-daybreak-blue-latest":{"mode":"responses","base_model":"gpt-daybreak-blue","max_input_tokens":922000,"max_output_tokens":128000,"fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"source":"https://developers.openai.com/api/docs/models/gpt-daybreak-blue-latest","supports_parallel_function_calling":true,"provider":"openai","web_search_billing_unit":"per_request","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"groq/llama-guard-3-8b":{"mode":"chat","base_model":"llama-guard-3-8b","max_input_tokens":8192,"max_tokens":8192,"source":"https://console.groq.com/docs/model/llama-guard-3-8b","provider":"groq","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false},"groq/meta-llama/llama-prompt-guard-2-22m":{"mode":"chat","base_model":"llama-prompt-guard-2-22m","max_input_tokens":512,"max_output_tokens":512,"source":"https://console.groq.com/docs/models","provider":"groq","max_tokens":512,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"groq/meta-llama/llama-prompt-guard-2-86m":{"mode":"chat","base_model":"llama-prompt-guard-2-86m","max_input_tokens":512,"max_output_tokens":512,"source":"https://console.groq.com/docs/model/meta-llama/llama-prompt-guard-2-86m","provider":"groq","max_tokens":512,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":512,"range":{"min":1,"max":512},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"groq/openai/gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":65536,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_web_search":true,"provider":"groq","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":false,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":false,"provider":"crusoe","max_tokens":163840,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"crusoe","max_tokens":163840,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/google/gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"crusoe","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"crusoe","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/moonshotai/Kimi-K2-Thinking":{"mode":"chat","base_model":"kimi-k2-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":false,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":false,"provider":"crusoe","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"crusoe","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"crusoe/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"crusoe","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"inception/mercury-2":{"mode":"chat","base_model":"mercury-2","max_input_tokens":128000,"max_output_tokens":50000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"inception","max_tokens":50000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":50000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"inception/mercury-2.5":{"mode":"chat","base_model":"mercury-2.5","max_input_tokens":260000,"max_output_tokens":65536,"source":"https://docs.inceptionlabs.ai/get-started/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"inception","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"text-completion-inception/mercury-edit-2":{"mode":"completion","base_model":"mercury-edit-2","max_input_tokens":32000,"max_output_tokens":8192,"provider":"text-completion-inception","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta/muse-spark-1.1":{"mode":"chat","base_model":"muse-spark-1.1","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://ai.developer.meta.com/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta/muse-spark-1.2":{"mode":"chat","base_model":"muse-spark-1.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://ai.developer.meta.com/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta/muse-spark-1.2-contributor":{"mode":"chat","base_model":"muse-spark-1.2-contributor","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://ai.developer.meta.com/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta/muse-spark-1.3":{"mode":"chat","base_model":"muse-spark-1.3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://ai.developer.meta.com/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"meta/muse-spark-1.3-contributor":{"mode":"chat","base_model":"muse-spark-1.3-contributor","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://ai.developer.meta.com/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":196000,"max_output_tokens":8000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"minimax/MiniMax-M3":{"mode":"chat","base_model":"minimax-m3","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"max_input_tokens":1000000,"max_output_tokens":128000,"provider":"minimax","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral.devstral-2-123b":{"mode":"chat","base_model":"devstral-2-123b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":256000,"max_output_tokens":32000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/devstral-small-latest":{"mode":"chat","base_model":"devstral-small","max_input_tokens":256000,"max_output_tokens":256000,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/devstral-small-2-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/devstral-latest":{"mode":"chat","base_model":"devstral","max_input_tokens":256000,"max_output_tokens":256000,"reasoning_effort_levels":["none","high"],"source":"https://mistral.ai/news/devstral-2-vibe-cli","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/devstral-medium-latest":{"mode":"chat","base_model":"devstral-medium","max_input_tokens":256000,"max_output_tokens":256000,"reasoning_effort_levels":["none","high"],"source":"https://mistral.ai/news/devstral-2-vibe-cli","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-14b-2512":{"mode":"chat","base_model":"ministral-14b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.mistral.ai/models/ministral-3-14b-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-14b-latest":{"mode":"chat","base_model":"ministral-14b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.mistral.ai/models/ministral-3-14b-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-3b-2512":{"mode":"chat","base_model":"ministral-3b","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://docs.mistral.ai/models/ministral-3-3b-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-medium-3":{"mode":"chat","base_model":"mistral-medium-3","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/voxtral-small-2507":{"mode":"chat","base_model":"voxtral-small","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://docs.mistral.ai/models/voxtral-small-25-07","supports_audio_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/voxtral-small-latest":{"mode":"chat","base_model":"voxtral-small","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://docs.mistral.ai/models/voxtral-small-25-07","supports_audio_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/zai-glm-5-2":{"mode":"chat","base_model":"glm-5-2","deprecation_date":"2026-10-31","deprecation_source":"https://docs.mistral.ai/getting-started/models/models_overview","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["none","minimal","low","medium","high","xhigh","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-2","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/zai-glm-5-3":{"mode":"chat","base_model":"glm-5-3","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-3","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/zai-glm-5":{"mode":"chat","base_model":"glm-5","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-3","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/zai-glm-latest":{"mode":"chat","base_model":"glm","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-3","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/glm-5-2":{"mode":"chat","base_model":"glm-5-2","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["none","minimal","low","medium","high","xhigh","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-2","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-large-2512":{"mode":"chat","base_model":"mistral-large","deprecation_date":"2027-04-30","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.mistral.ai/models/mistral-large-3-25-12","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-medium-2604":{"mode":"chat","base_model":"mistral-medium","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-medium-3-5":{"mode":"chat","base_model":"mistral-medium-3-5","deprecation_date":"2027-04-30","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-3-3b-2512":{"mode":"chat","base_model":"ministral-3-3b","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://mistral.ai/pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-3-8b-2512":{"mode":"chat","base_model":"ministral-3-8b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://mistral.ai/pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-3-14b-2512":{"mode":"chat","base_model":"ministral-3-14b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://mistral.ai/pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/ministral-8b-2512":{"mode":"chat","base_model":"ministral-8b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://mistral.ai/pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"moonshot/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://platform.kimi.ai/docs/pricing/chat-k27-code","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"moonshot","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"model":"kimi-k2.7-code","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"moonshot/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://platform.kimi.ai/docs/pricing/chat-k26","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"moonshot","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_prompt_caching":true,"supports_web_search":true,"model":"kimi-k2.6","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"moonshot/kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"reasoning_effort_levels":["low","high","max"],"source":"https://platform.kimi.ai/docs/pricing/chat-k3","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"moonshot","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_web_search":true,"model":"kimi-k3","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":164000,"max_output_tokens":164000,"supports_function_calling":true,"supports_reasoning":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":164000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":164000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"deepseek-r1-distill-llama-70b","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/google/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","max_input_tokens":110000,"max_output_tokens":110000,"supports_function_calling":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/google%2Fgemma-3-27b-it","provider":"nebius","max_tokens":110000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":110000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/meta-llama/Llama-Guard-3-8B":{"mode":"chat","base_model":"llama-guard-3-8b","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/meta-llama/Meta-Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/meta-llama/Meta-Llama-3.1-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/meta-llama/Meta-Llama-3.1-405B-Instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/mistralai/Mistral-Nemo-Instruct-2407":{"mode":"chat","base_model":"mistral-nemo-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/NousResearch/Hermes-3-Llama-3.1-405B":{"mode":"chat","base_model":"hermes-3-llama-3.1-405b","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Llama-3.1-Nemotron-Ultra-253B-v1":{"mode":"chat","base_model":"llama-3.1-nemotron-ultra-253b-v1","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Llama-3.3-Nemotron-Super-49B-v1":{"mode":"chat","base_model":"llama-3.3-nemotron-super-49b-v1","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-235B-A22B":{"mode":"chat","base_model":"qwen3-235b-a22b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-32B":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":40960,"max_output_tokens":40960,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/Qwen%2FQwen3-32B","provider":"nebius","max_tokens":40960,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-30B-A3B":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-14B":{"mode":"chat","base_model":"qwen3-14b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-4B":{"mode":"chat","base_model":"qwen3-4b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/QwQ-32B":{"mode":"chat","base_model":"qwq-32b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_reasoning":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-72b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2.5-32B-Instruct":{"mode":"chat","base_model":"qwen2.5-32b-instruct","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2.5-Coder-7B":{"mode":"chat","base_model":"qwen2.5-coder-7b","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2.5-VL-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-vl-72b-instruct","max_input_tokens":32000,"max_output_tokens":32000,"supports_function_calling":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/image2text/Qwen%2FQwen2.5-VL-72B-Instruct","provider":"nebius","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2-VL-72B-Instruct":{"mode":"chat","base_model":"qwen2-vl-72b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_vision":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen2-VL-7B-Instruct":{"mode":"chat","base_model":"qwen2-vl-7b-instruct","max_input_tokens":131072,"max_output_tokens":131072,"supports_vision":true,"source":"https://nebius.com/prices","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/deepseek-ai%2FDeepSeek-V4-Flash","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1024000,"max_output_tokens":1024000,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/deepseek-ai%2FDeepSeek-V4-Flash-0731","provider":"nebius","max_tokens":1024000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1024000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/deepseek-ai%2FDeepSeek-V4-Pro","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/deepseek-ai%2FDeepSeek-V4-Pro-0813","supports_function_calling":true,"supports_reasoning":true,"provider":"nebius","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/MiniMaxAI/MiniMax-M2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":196608,"max_output_tokens":196608,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/MiniMaxAI%2FMiniMax-M2.5","provider":"nebius","max_tokens":196608,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/MiniMaxAI%2FMiniMax-M3","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/image2text/moonshotai%2FKimi-K2.6","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/moonshotai%2FKimi-K2.7-Code","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/moonshotai/Kimi-K3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1024000,"max_output_tokens":1024000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/image2text/moonshotai%2FKimi-K3","provider":"nebius","max_tokens":1024000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1024000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/NousResearch/Hermes-4-405B":{"mode":"chat","base_model":"hermes-4-405b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/NousResearch%2FHermes-4-405B","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/NousResearch/Hermes-4-70B":{"mode":"chat","base_model":"hermes-4-70b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/NousResearch%2FHermes-4-70B","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Cosmos3-Super-Reasoner":{"mode":"chat","base_model":"cosmos3-super-reasoner","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/image2text/nvidia%2FCosmos3-Super-Reasoner","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1":{"mode":"chat","base_model":"llama-3-1-nemotron-ultra-253b-v1","max_input_tokens":131072,"max_output_tokens":131072,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2FLlama-3_1-Nemotron-Ultra-253B-v1","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B":{"mode":"chat","base_model":"nvidia-nemotron-3-nano-30b-a3b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2FNVIDIA-Nemotron-3-Nano-30B-A3B","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Nemotron-3-Nano-Omni":{"mode":"chat","base_model":"nemotron-3-nano-omni","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2FNemotron-3-Nano-Omni","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/nemotron-3-super-120b-a12b":{"mode":"chat","base_model":"nemotron-3-super-120b-a12b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2Fnemotron-3-super-120b-a12b","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Nemotron-3-Ultra-550b-a55b":{"mode":"chat","base_model":"nemotron-3-ultra-550b-a55b","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2FNemotron-3-Ultra-550b-a55b","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/nvidia/Nemotron-3_5-Lightning":{"mode":"chat","base_model":"nemotron-3-5-lightning","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/nvidia%2FNemotron-3_5-Lightning","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/openai%2Fgpt-oss-120b","provider":"nebius","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/openbmb/MiniCPM-V-4_5":{"mode":"chat","base_model":"openbmb/minicpm-v-4-5","max_input_tokens":32000,"max_output_tokens":32000,"supports_vision":true,"source":"https://tokenfactory.nebius.com/models/catalog/image2text/openbmb%2FMiniCPM-V-4_5","provider":"nebius","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/Qwen%2FQwen3-235B-A22B-Instruct-2507","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-30B-A3B-Instruct-2507":{"mode":"chat","base_model":"qwen3-30b-a3b-instruct","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/Qwen%2FQwen3-30B-A3B-Instruct-2507","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3-Next-80B-A3B-Thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/Qwen%2FQwen3-Next-80B-A3B-Thinking","provider":"nebius","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"qwen3.5-397b-a17b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/Qwen%2FQwen3.5-397B-A17B","provider":"nebius","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/zai-org/GLM-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":202752,"max_output_tokens":202752,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/zai-org%2FGLM-5.1","provider":"nebius","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/zai-org%2FGLM-5.2","provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/zai-org/GLM-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":1024000,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/zai-org%2FGLM-5.3","supports_function_calling":true,"supports_reasoning":true,"provider":"nebius","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nebius/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1024000,"max_output_tokens":1024000,"supports_function_calling":true,"supports_reasoning":true,"source":"https://tokenfactory.nebius.com/models/catalog/text2text/zai-org%2FGLM-5.3-Flash","supports_vision":true,"provider":"nebius","max_tokens":1024000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1024000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nvidia-nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","max_tokens":32768,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/meta.llama-3.1-8b-instruct":{"mode":"chat","base_model":"llama-3.1-8b-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/meta.llama-3.1-70b-instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/google.gemini-2.5-flash":{"mode":"chat","base_model":"gemini-2.5-flash","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_native_streaming":true,"supports_image_size":false,"supports_reasoning":true,"supports_system_messages":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"provider":"oci","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/google.gemini-2.5-pro":{"mode":"chat","base_model":"gemini-2.5-pro","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_system_messages":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"provider":"oci","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/google.gemini-2.5-flash-lite":{"mode":"chat","base_model":"gemini-2.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_vision":true,"supports_native_streaming":true,"supports_image_size":false,"supports_reasoning":false,"supports_system_messages":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"provider":"oci","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-a-vision":{"mode":"chat","base_model":"command-a-vision","max_input_tokens":256000,"max_output_tokens":8192,"source":"https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/","supports_function_calling":true,"supports_response_schema":false,"supports_native_streaming":true,"supports_vision":true,"provider":"oci","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-a-reasoning":{"mode":"chat","base_model":"command-a","max_input_tokens":256000,"max_output_tokens":8192,"source":"https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/","supports_function_calling":false,"supports_response_schema":false,"supports_native_streaming":true,"provider":"oci","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-a-reasoning-08-2025":{"mode":"chat","base_model":"command-a","max_input_tokens":256000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-a-vision-07-2025":{"mode":"chat","base_model":"command-a-vision","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_vision":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-a-translate-08-2025":{"mode":"chat","base_model":"command-a-translate","max_input_tokens":256000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":false,"supports_response_schema":false,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-r-08-2024":{"mode":"chat","base_model":"command-r","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/cohere.command-r-plus-08-2024":{"mode":"chat","base_model":"command-r-plus","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/meta.llama-3.2-11b-vision-instruct":{"mode":"chat","base_model":"llama-3.2-11b-vision-instruct","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"supports_vision":true,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/meta.llama-3.3-70b-instruct-fp8-dynamic":{"mode":"chat","base_model":"llama-3.3-70b-instruct-fp8-dynamic","max_input_tokens":128000,"max_output_tokens":4000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/xai.grok-4-fast":{"mode":"chat","base_model":"grok-4-fast","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/xai.grok-4.1-fast":{"mode":"chat","base_model":"grok-4.1-fast","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/xai.grok-4.20":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":131072,"max_output_tokens":131072,"reasoning_effort_levels":[],"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/xai.grok-4.20-multi-agent":{"mode":"chat","base_model":"grok-4.20-multi-agent","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/xai.grok-code-fast-1":{"mode":"chat","base_model":"grok-code-fast-1","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_response_schema":false,"provider":"oci","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/openai.gpt-5":{"mode":"chat","base_model":"gpt-5","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"oci","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/openai.gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"oci","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"oci/openai.gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"oci","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-sonnet-4.6":{"mode":"chat","base_model":"claude-sonnet-4.6","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_adaptive_thinking":true,"supports_legacy_thinking":true,"source":"https://openrouter.ai/api/v1/models","supports_assistant_prefill":true,"supports_computer_use":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.6":{"mode":"chat","base_model":"claude-opus-4.6","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":4096,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.7":{"mode":"chat","base_model":"claude-opus-4.7","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_max_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"prompt_cache_min_tokens":2048,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"source":"https://openrouter.ai/api/v1/models","supports_assistant_prefill":false,"supports_audio_input":false,"supports_computer_use":true,"supports_max_reasoning_effort":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-5.5":{"mode":"chat","base_model":"claude-opus-5.5","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","provider":"openrouter","max_input_tokens":1024000,"max_output_tokens":384000,"max_tokens":384000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4.1-flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"off_peak_pricing":{"windows":[{"weekdays":["saturday","sunday"],"hours_utc":"00:00-00:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"00:00-01:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"04:00-06:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"10:00-00:00"}],"input_cost_per_token":1.5e-7,"output_cost_per_token":6e-7,"cache_read_input_token_cost":3e-9},"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro-0813","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":384000,"max_tokens":384000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","off_peak_pricing":{"windows":[{"weekdays":["saturday","sunday"],"hours_utc":"00:00-00:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"00:00-01:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"04:00-06:00"},{"weekdays":["monday","tuesday","wednesday","thursday","friday"],"hours_utc":"10:00-00:00"}],"input_cost_per_token":6.6e-7,"output_cost_per_token":0.00000198,"cache_read_input_token_cost":2.2e-8},"supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-lite-preview":{"mode":"chat","base_model":"gemini-3.1-flash-lite-preview","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"rpm":2000,"source":"https://openrouter.ai/api/v1/models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_url_context":true,"supports_video_input":true,"supports_web_search":true,"tpm":800000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"rpm":2000,"source":"https://openrouter.ai/api/v1/models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_url_context":true,"supports_video_input":true,"supports_web_search":true,"tpm":800000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-pro-preview":{"mode":"chat","base_model":"gemini-3.1-pro-preview","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_system_messages":true,"supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3.5-lightning":{"mode":"chat","base_model":"nemotron-3.5-lightning","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.1-codex-max":{"mode":"chat","base_model":"gpt-5.1-codex-max","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://openrouter.ai/api/v1/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-sol-pro":{"mode":"chat","base_model":"gpt-5.6-sol-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-coder-plus":{"mode":"chat","base_model":"qwen3-coder-plus","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.6-plus":{"mode":"chat","base_model":"qwen3.6-plus","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-35b-a3b":{"mode":"chat","base_model":"qwen3.5-35b-a3b","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-27b":{"mode":"chat","base_model":"qwen3.5-27b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-122b-a10b":{"mode":"chat","base_model":"qwen3.5-122b-a10b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-flash-02-23":{"mode":"chat","base_model":"qwen3.5-flash-02-23","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-plus-02-15":{"mode":"chat","base_model":"qwen3.5-plus-02-15","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-397b-a17b":{"mode":"chat","base_model":"qwen3.5-397b-a17b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/xiaomi/mimo-v2.5-pro":{"mode":"chat","base_model":"mimo-v2.5-pro","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/xiaomi/mimo-v2.5":{"mode":"chat","base_model":"mimo-v2.5","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_video_input":true,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5":{"mode":"chat","base_model":"glm-5","provider":"openrouter","max_input_tokens":198000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.1":{"mode":"chat","base_model":"glm-5.1","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_vision":false,"supports_computer_use":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openrouter/free":{"mode":"chat","base_model":"free","provider":"openrouter","max_input_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openrouter/bodybuilder":{"mode":"chat","base_model":"bodybuilder","max_input_tokens":128000,"max_tokens":128000,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"perplexity/preset/fast-search":{"mode":"responses","base_model":"preset/fast-search","supports_web_search":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/preset/deep-research":{"mode":"responses","base_model":"preset/deep-research","supports_web_search":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/preset/advanced-deep-research":{"mode":"responses","base_model":"preset/advanced","supports_web_search":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.1":{"mode":"responses","base_model":"gpt-5.1","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5-mini":{"mode":"responses","base_model":"gpt-5-mini","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-opus-4-6":{"mode":"responses","base_model":"claude-opus-4-6","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"supports_output_config":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-opus-4-7":{"mode":"responses","base_model":"claude-opus-4-7","supports_adaptive_thinking":true,"supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"supports_output_config":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-opus-4-5":{"mode":"responses","base_model":"claude-opus-4-5","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"supports_output_config":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-sonnet-4-5":{"mode":"responses","base_model":"claude-sonnet-4-5","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-haiku-4-5":{"mode":"responses","base_model":"claude-haiku-4-5","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3-pro-preview":{"mode":"responses","base_model":"gemini-3-pro","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3-flash-preview":{"mode":"responses","base_model":"gemini-3-flash","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-2.5-pro":{"mode":"responses","base_model":"gemini-2.5-pro","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-2.5-flash":{"mode":"responses","base_model":"gemini-2.5-flash","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"supports_image_size":false,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4-1-fast-non-reasoning":{"mode":"responses","base_model":"grok-4-1-fast-non","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/sonar":{"mode":"chat","base_model":"sonar","max_input_tokens":128000,"max_output_tokens":128000,"supports_web_search":true,"provider":"perplexity","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"perplexity/perplexity/deepseek-v4-flash-0731":{"mode":"responses","base_model":"deepseek-v4-flash","source":"https://docs.perplexity.ai/docs/agent-api/models","supports_web_search":true,"supports_reasoning":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/glm-5.2":{"mode":"responses","base_model":"glm-5.2","source":"https://docs.perplexity.ai/docs/agent-api/models","supports_web_search":true,"supports_reasoning":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/kimi-k3":{"mode":"responses","base_model":"kimi-k3","reasoning_effort_levels":["minimal","low","medium","high","xhigh","max"],"source":"https://docs.perplexity.ai/docs/agent-api/models","supports_web_search":true,"supports_reasoning":true,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/kimi-k2.7-code":{"mode":"responses","base_model":"kimi-k2.7-code","source":"https://docs.perplexity.ai/docs/agent-api/models","supports_web_search":true,"supports_reasoning":false,"supports_function_calling":true,"provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"bedrock/ap-northeast-1/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/ap-south-1/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/ap-southeast-2/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-west-1/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/eu-west-2/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/sa-east-1/qwen.qwen3-next-80b-a3b":{"mode":"chat","base_model":"qwen3-next-80b-a3b","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"replicate/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","source":"https://replicate.com/openai/gpt-oss-20b","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"sambanova/MiniMax-M2.7":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":196608,"max_output_tokens":131072,"source":"https://cloud.sambanova.ai/plans/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"sambanova","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"sambanova/DeepSeek-V3.2":{"mode":"chat","base_model":"deepseek-v3.2","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"sambanova/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":131072,"max_output_tokens":131072,"supports_vision":true,"source":"https://cloud.sambanova.ai/plans/pricing","provider":"sambanova","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scx-ai/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://scx.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"scx-ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scx-ai/Qwen3.8-Max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://scx.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"scx-ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"minimax-m3","deprecation_date":"2026-11-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":524288,"max_output_tokens":524288,"source":"https://api.together.ai/v1/models","successor_model":"together_ai/MiniMaxAI/MiniMax-M3.1-Flash","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","is_deprecated":true,"max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Prism-ML/Ternary-Bonsai-27B":{"mode":"chat","base_model":"prism-ml/ternary-bonsai-27b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://docs.together.ai/docs/serverless-models","provider":"together_ai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.5-9B":{"mode":"chat","base_model":"qwen3.5-9b","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.6-Plus":{"mode":"chat","base_model":"qwen3.6-plus","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.together.ai/v1/models","supports_reasoning":true,"provider":"together_ai","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.7-Max":{"mode":"chat","base_model":"qwen3.7-max","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.together.ai/v1/models","supports_prompt_caching":true,"provider":"together_ai","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.7-Plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.together.ai/v1/models","provider":"together_ai","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.8-2.4T-A95B":{"mode":"chat","base_model":"qwen3.8-2.4t-a95b","deprecation_date":"2026-11-04","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":1010000,"max_output_tokens":1010000,"source":"https://api.together.ai/v1/models","successor_model":"together_ai/Qwen/Qwen3.7-Max","supports_prompt_caching":true,"provider":"together_ai","is_deprecated":true,"max_tokens":1010000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1010000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/arize-ai/qwen-2-1.5b-instruct":{"mode":"chat","base_model":"arize-ai/qwen-2-1.5b-instruct","max_input_tokens":32768,"max_output_tokens":32768,"source":"https://api.together.ai/v1/models","provider":"together_ai","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-09-29","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.together.ai/v1/models","successor_model":"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"deprecation_date":"2026-09-29","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/meta-models/Muse-Glimmer-30B":{"mode":"chat","base_model":"models/muse-glimmer-30b","max_input_tokens":131072,"max_output_tokens":131072,"source":"https://api.together.ai/v1/models","supports_prompt_caching":true,"provider":"together_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/moonshotai/Kimi-K3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"reasoning_effort_levels":["low","high","max"],"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/thinkingmachines/Inkling":{"mode":"chat","base_model":"thinkingmachines/inkling","deprecation_date":"2026-11-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":524288,"max_output_tokens":524288,"source":"https://api.together.ai/v1/models","successor_model":"together_ai/meta-models/Muse-Spark-Open-816B-A42B","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","is_deprecated":true,"max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048575,"max_output_tokens":128000,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/zai-org/GLM-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048575,"max_output_tokens":128000,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"together_ai","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048575,"max_output_tokens":128000,"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"supports_forced_tool_use":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-fable-5-1","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nvidia-nemotron-nano-3-30b","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-12b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html","supports_system_messages":true,"supports_vision":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-9b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html","max_input_tokens":128000,"max_output_tokens":8192,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html","supports_system_messages":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nvidia-nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":500000,"max_output_tokens":500000,"supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","max_tokens":500000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","deprecation_date":"2026-10-15","successor_model":"vertex_ai/claude-haiku-5-5","max_input_tokens":200000,"max_output_tokens":64000,"regional_endpoint_uplift_multiplier":1.1,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_native_streaming":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"provider":"vertex_ai","supports_web_search":true,"is_deprecated":true,"max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-4-6@default":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"vertex_ai","supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","deprecation_date":"2027-06-08","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-fable-5@default":{"mode":"chat","base_model":"claude-fable-5","deprecation_date":"2027-06-08","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","deprecation_date":"2027-01-24","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"vertex_multi_region_only":true,"supports_web_search":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-5@default":{"mode":"chat","base_model":"claude-opus-5","deprecation_date":"2027-01-24","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_safeguards":true,"supports_web_search":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","provider_specific_entry":{"us":1.1},"supports_output_config":true,"tool_use_system_prompt_tokens":286,"supports_web_search":true,"supports_forced_tool_choice":false,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-5-5@default":{"mode":"chat","base_model":"claude-opus-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_safeguards":true,"supports_web_search":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","deprecation_date":"2027-05-28","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","max_tokens":1000000,"tool_use_system_prompt_tokens":346,"supports_minimal_reasoning_effort":true,"vertex_multi_region_only":true,"supports_web_search":true,"supports_safeguards":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-opus-4-8@default":{"mode":"chat","base_model":"claude-opus-4-8","deprecation_date":"2027-05-28","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"supports_adaptive_thinking":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_web_search":true,"supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","deprecation_date":"2026-12-24","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_safeguards":true,"supports_web_search":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"vertex_ai","supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3-pro-image":{"mode":"image_generation","base_model":"gemini-3-pro-image","max_input_tokens":65536,"max_output_tokens":32768,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supports_reasoning":false,"provider":"vertex_ai","deprecation_date":"2027-05-28","max_tokens":32768,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.1-flash-image":{"mode":"image_generation","base_model":"gemini-3.1-flash-image","max_input_tokens":65536,"max_output_tokens":32768,"regional_endpoint_uplift_multiplier":1.1,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supports_reasoning":false,"provider":"vertex_ai","deprecation_date":"2027-05-28","max_tokens":32768,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.1-flash-image-preview":{"mode":"image_generation","base_model":"gemini-3.1-flash-image","max_input_tokens":65536,"max_output_tokens":32768,"supports_reasoning":false,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"provider":"vertex_ai","max_tokens":32768,"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.1-flash-lite-image":{"mode":"image_generation","base_model":"gemini-3.1-flash-lite-image","max_input_tokens":65536,"max_output_tokens":4096,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":4096,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-3.1-flash-lite-preview":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/google/gemma-4-26b-a4b-it-maas":{"mode":"chat","base_model":"gemma-4","max_input_tokens":262144,"max_output_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supported_regions":["global"],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":8192,"tpm":250000,"rpm":10,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://wandb.ai/inference/coreweave/cw_moonshotai_Kimi-K2.5","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"watsonx/bigscience/mt0-xxl":{"mode":"chat","base_model":"mt0-xxl","max_input_tokens":4096,"max_output_tokens":4096,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":4096,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"watsonx/meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"mode":"chat","base_model":"llama-4-maverick-17b-128e-instruct-fp8","max_input_tokens":131072,"max_output_tokens":8192,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":false,"source":"https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/fm-models.html?context=wx","provider":"watsonx","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-multi-agent-beta-0309":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supported_endpoints":["/v1/responses"],"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-4.20-beta-0309-reasoning":{"mode":"chat","base_model":"grok-4.20-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-0309-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-0309-non-reasoning":{"mode":"chat","base_model":"grok-4.20-beta-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.3":{"mode":"chat","base_model":"grok-4.3","token_count_framing":185,"token_count_tool_framing":214,"max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.3-latest":{"mode":"chat","base_model":"grok-4.3","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.5":{"mode":"chat","base_model":"grok-4.5","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://docs.x.ai/developers/models/grok-4.5","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":500000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.5-latest":{"mode":"chat","base_model":"grok-4.5","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://docs.x.ai/developers/models/grok-4.5","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":500000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-build-latest":{"mode":"chat","base_model":"grok-build","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":500000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://docs.x.ai/developers/models/grok-4.6","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":500000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.7":{"mode":"chat","base_model":"grok-4.7","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://docs.x.ai/developers/models/grok-4.7","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"xai","max_tokens":500000,"supports_audio_input":false,"supports_audio_output":false,"regional_endpoint_uplift_multiplier":1.1,"web_search_billing_unit":"per_query","default_reasoning_effort":"high","reasoning_effort_levels":["low","medium","high","xhigh"],"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"thinking_always_on":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_code_execution":true,"supports_service_tier":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_regions":["us-east-1","us-west-2","us-central-1"],"rpm":9000,"tpm":50000000,"metadata":{"notes":"Released 2026-09-21. Knowledge cutoff May 2026. Long-context rates apply to all tokens in a request once the prompt reaches 200k tokens (>= 200k, not strictly above). No documented text output limit; max_output_tokens mirrors the 500k context window. Batch API not supported. rpm derived from 150 requests/second. US regional endpoint (us.api.x.ai) billed at 1.1x. Grok 4.7 Fast (2x rates) is Cursor/Grok Build only and not on the public API, so it has no entry.","calculation":"Priority tier = 2x standard rates on all token types, applied after caching discount. Web search = $5 per 1k calls = $0.005 per call."},"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai.glm-5":{"mode":"chat","base_model":"zai.glm-5","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai.glm-4.7-flash":{"mode":"chat","base_model":"zai.glm-4.7-flash","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":203000,"max_output_tokens":4000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5":{"mode":"chat","base_model":"glm-5","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://docs.z.ai/guides/overview/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5.3-flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","supports_vision":true,"provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5-code":{"mode":"chat","base_model":"glm-5-code","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-4.7-flash":{"mode":"chat","base_model":"glm-4.7-flash","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"source":"https://docs.z.ai/guides/overview/pricing","provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/qwen3p7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/glm-5p1-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/glm-5.1-fast","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/kimi-k2p6-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/kimi-k2.6-fast","deprecation_date":"2026-08-27","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/accounts/fireworks/models/kimi-k2p6","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/kimi-k2p7-code-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/kimi-k2.7-code-fast","deprecation_date":"2026-08-27","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/accounts/fireworks/models/kimi-k2p7-code","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/qwen/qwen3.5-397b-a17b":{"mode":"chat","base_model":"qwen3.5-397b-a17b","max_input_tokens":256000,"max_output_tokens":16384,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"scaleway","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/qwen/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_input_tokens":256000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_reasoning":true,"provider":"scaleway","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/qwen/qwen3-235b-a22b-instruct-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-instruct","max_input_tokens":256000,"max_output_tokens":16384,"supports_function_calling":true,"provider":"scaleway","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/qwen/qwen3-coder-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","max_input_tokens":128000,"max_output_tokens":32768,"supports_function_calling":true,"deprecation_date":"2026-10-01","provider":"scaleway","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128000,"max_output_tokens":32768,"supports_function_calling":true,"provider":"scaleway","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","max_input_tokens":256000,"max_output_tokens":32768,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"scaleway","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/mistralai/mistral-medium-3.5-128b":{"mode":"chat","base_model":"mistral-medium-3.5-128b","max_input_tokens":256000,"max_output_tokens":16384,"supports_reasoning":true,"supports_function_calling":true,"supports_vision":true,"supports_tool_choice":true,"provider":"scaleway","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/mistralai/mistral-small-3.2-24b-instruct-2506":{"mode":"chat","base_model":"mistral-small-3.2-24b-instruct","max_input_tokens":128000,"max_output_tokens":32768,"supports_function_calling":true,"supports_vision":true,"provider":"scaleway","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/mistralai/pixtral-12b-2409":{"mode":"chat","base_model":"pixtral-12b","max_input_tokens":128000,"max_output_tokens":4096,"supports_vision":true,"supports_function_calling":true,"deprecation_date":"2026-10-01","provider":"scaleway","is_deprecated":true,"max_tokens":4096,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/meta/llama-3.3-70b-instruct":{"mode":"chat","base_model":"llama-3.3-70b-instruct","max_input_tokens":128000,"max_output_tokens":16384,"supports_function_calling":true,"provider":"scaleway","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/hermes-3-8b-tee":{"mode":"chat","base_model":"hermes-3-8b-tee","max_input_tokens":16000,"max_output_tokens":16000,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":false,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/gemma-4-31b-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/gemma-4-31b-it-thinking":{"mode":"chat","base_model":"gemma-4-31b-it-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.6-27b":{"mode":"chat","base_model":"qwen3.6-27b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.6-27b-thinking":{"mode":"chat","base_model":"qwen3.6-27b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.6-35b-a3b-thinking":{"mode":"chat","base_model":"qwen3.6-35b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.5-122b-a10b":{"mode":"chat","base_model":"qwen3.5-122b-a10b","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/qwen3.5-122b-a10b-thinking":{"mode":"chat","base_model":"qwen3.5-122b-a10b-thinking","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":200000,"max_output_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":false,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":200000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"libertai/deepseek-v4-flash-thinking":{"mode":"chat","base_model":"deepseek-v4-flash-thinking","max_input_tokens":200000,"max_output_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_vision":false,"supports_reasoning":true,"source":"https://docs.libertai.io/apis/text/","provider":"libertai","max_tokens":200000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-sonnet-5@default":{"mode":"chat","base_model":"claude-sonnet-5","deprecation_date":"2026-12-24","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":1024,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_safeguards":true,"supports_web_search":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/claude-sonnet-4-6@default":{"mode":"chat","base_model":"claude-sonnet-4-6","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"vertex_ai","supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-oss-safeguard-120b":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":65536,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":65536,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-5.6-sol":{"mode":"responses","base_model":"gpt-5.6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/openai.gpt-5.6-terra":{"mode":"responses","base_model":"gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/openai.gpt-5.6-cyber":{"mode":"responses","base_model":"gpt-5.6-cyber","max_input_tokens":272000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-cyber.html","use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_web_search":true,"provider":"bedrock_mantle","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/openai.gpt-daybreak-blue-5.6-sol":{"mode":"responses","base_model":"gpt-daybreak-blue-5.6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-daybreak-blue-56-sol.html","supports_web_search":true,"provider":"bedrock_mantle","bedrock_mantle_base_path":"openai/v1","max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/openai.gpt-5.6-luna":{"mode":"responses","base_model":"gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"us.openai.gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"global.openai.gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us.openai.gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"global.openai.gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us.openai.gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"global.openai.gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/openai.gpt-6-astra":{"mode":"responses","base_model":"gpt-6-astra","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-astra.html","supports_web_search":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-east-1","us-east-2","us-west-1","ca-central-1","eu-central-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":128000,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"supports_service_tier":true,"service_tiers":["auto","default","ultrafast"],"bedrock_mantle_base_path":"openai/v1","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.openai.gpt-6-astra":{"mode":"responses","base_model":"gpt-6-astra","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-astra.html","supported_endpoints":["/v1/responses"],"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-west-2","us-east-1","us-east-2","us-west-1","ca-central-1","eu-central-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_tokens":128000,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default","ultrafast"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/openai.gpt-5.5":{"mode":"responses","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/openai.gpt-5.4":{"mode":"responses","base_model":"gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-west-1","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","us-gov-east-1"],"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/google.gemma-4-31b":{"mode":"chat","base_model":"gemma-4-31b","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/google.gemma-4-26b-a4b":{"mode":"chat","base_model":"gemma-4-26b-a4b","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/google.gemma-4-e2b":{"mode":"chat","base_model":"gemma-4-e2b","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":true,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/anthropic.claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-0-pro-260215":{"mode":"chat","base_model":"doubao-seed-2-0-pro","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://docs.volcengine.com/docs/82379/1330310","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":4.6e-7,"output_cost_per_token":0.0000023,"range":[0,32000]},{"input_cost_per_token":7e-7,"output_cost_per_token":0.0000035,"range":[32000,128000]},{"input_cost_per_token":0.0000014,"output_cost_per_token":0.000007,"range":[128000,256000]}],"provider":"volcengine","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-1-pro-260628":{"mode":"chat","base_model":"doubao-seed-2-1-pro","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://www.volcengine.com/docs/82379/1544106","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"provider":"volcengine","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-1-turbo-260628":{"mode":"chat","base_model":"doubao-seed-2-1-turbo","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://www.volcengine.com/docs/82379/1544106","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"provider":"volcengine","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-0-lite-260215":{"mode":"chat","base_model":"doubao-seed-2-0-lite","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://docs.volcengine.com/docs/82379/1330310","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":8.7e-8,"output_cost_per_token":5.2e-7,"range":[0,32000]},{"input_cost_per_token":1.3e-7,"output_cost_per_token":7.8e-7,"range":[32000,128000]},{"input_cost_per_token":2.6e-7,"output_cost_per_token":0.0000016,"range":[128000,256000]}],"provider":"volcengine","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-0-mini-260215":{"mode":"chat","base_model":"doubao-seed-2-0-mini","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://docs.volcengine.com/docs/82379/1330310","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":2.9e-8,"output_cost_per_token":2.9e-7,"range":[0,32000]},{"input_cost_per_token":5.8e-8,"output_cost_per_token":5.8e-7,"range":[32000,128000]},{"input_cost_per_token":1.2e-7,"output_cost_per_token":0.0000012,"range":[128000,256000]}],"provider":"volcengine","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"volcengine/doubao-seed-2-0-code-preview-260215":{"mode":"chat","base_model":"doubao-seed-2-0-code","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://docs.volcengine.com/docs/82379/1330310","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"tiered_pricing":[{"input_cost_per_token":4.6e-7,"output_cost_per_token":0.0000023,"range":[0,32000]},{"input_cost_per_token":7e-7,"output_cost_per_token":0.0000035,"range":[32000,128000]},{"input_cost_per_token":0.0000014,"output_cost_per_token":0.000007,"range":[128000,256000]}],"provider":"volcengine","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-east-1/zai.glm-5":{"mode":"chat","base_model":"zai.glm-5","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-west-2/zai.glm-5":{"mode":"chat","base_model":"zai.glm-5","max_input_tokens":200000,"max_output_tokens":128000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"prompt_cache_min_tokens":1024,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"prompt_cache_min_tokens":1024,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-4-sonnet":{"mode":"chat","base_model":"claude-sonnet-4","max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"prompt_cache_min_tokens":1024,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-4-opus":{"mode":"chat","base_model":"claude-opus-4","max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"prompt_cache_min_tokens":1024,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"prompt_cache_min_tokens":4096,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/claude-3-7-sonnet":{"mode":"chat","base_model":"claude-3-7-sonnet","max_input_tokens":200000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/openai-gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","max_input_tokens":300000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/openai-gpt-5":{"mode":"chat","base_model":"gpt-5","max_input_tokens":300000,"max_output_tokens":16384,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/openai-gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","max_input_tokens":1000000,"max_output_tokens":16384,"supports_function_calling":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/openai-gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","max_input_tokens":5000000,"max_output_tokens":16384,"supports_function_calling":true,"supports_system_messages":true,"supports_response_schema":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"snowflake/llama4-maverick":{"mode":"chat","base_model":"llama-4-maverick","max_input_tokens":128000,"max_output_tokens":16384,"supports_function_calling":true,"supports_system_messages":true,"provider":"snowflake","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/Qwen/Qwen3.5-397B-A17B-FP8":{"mode":"chat","base_model":"qwen3.5-397b-a17b-fp8","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct-fp8","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/Qwen/Qwen3.6-27B-FP8":{"mode":"chat","base_model":"qwen3.6-27b-fp8","max_input_tokens":262144,"max_output_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/lukealonso/GLM-5.1-NVFP4-MTP":{"mode":"chat","base_model":"lukealonso/glm-5.1-nvfp4-mtp","max_input_tokens":202752,"max_output_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/MiniMaxAI/MiniMax-M2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":196608,"max_output_tokens":196608,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":32768,"max_output_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tensormesh/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_system_messages":true,"supports_reasoning":true,"source":"https://serverless.tensormesh.ai/v1/models/openrouter","provider":"tensormesh","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepseek-flash":{"mode":"chat","base_model":"deepseek-flash","max_input_tokens":1048576,"max_output_tokens":393216,"off_peak_pricing":{"cache_read_input_token_cost":3e-9,"input_cost_per_token":1.5e-7,"output_cost_per_token":6e-7,"windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses","/v1/messages"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"deepseek","max_tokens":393216,"peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"supported_modalities":["text","image"],"supported_output_modalities":["text"],"metadata":{"notes":"DeepSeek-V4.1-Flash. Legacy names deepseek-v4-flash and deepseek-v4-flash-vision-exp route here at Flash pricing; deepseek-v4-pro also routes here from 2026-09-14 until V4.1 Pro ships.","calculation":"Peak: $0.30/M cache-miss input, $0.006/M cache-hit input, $1.20/M output. Off-peak = 0.5x."},"supports_native_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"off_peak_pricing":{"cache_read_input_token_cost":3e-9,"input_cost_per_token":1.5e-7,"output_cost_per_token":6e-7,"windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"token_count_framing":23,"token_count_tool_framing":12,"provider":"deepseek","max_tokens":393216,"peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"server_side_model":"deepseek-flash","supports_native_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision-exp","max_input_tokens":1000000,"max_output_tokens":393216,"off_peak_pricing":{"cache_read_input_token_cost":3e-9,"input_cost_per_token":1.5e-7,"output_cost_per_token":6e-7,"windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"deepseek","max_tokens":393216,"peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"supports_native_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":393216,"off_peak_pricing":{"cache_read_input_token_cost":2.2e-8,"input_cost_per_token":6.6e-7,"output_cost_per_token":0.00000198,"windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"token_count_framing":76,"token_count_tool_framing":75,"provider":"deepseek","max_tokens":393216,"peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"supports_native_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tencent/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":384000,"source":"https://www.tencentcloud.com/products/tokenhub","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"provider":"tencent","max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tencent/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":384000,"source":"https://www.tencentcloud.com/products/tokenhub","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"provider":"tencent","max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"tencent/minimax-m3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":1000000,"source":"https://www.tencentcloud.com/products/tokenhub","supported_endpoints":["/v1/chat/completions"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"tencent","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cognition/swe-1.6":{"mode":"chat","base_model":"swe-1.6","supports_function_calling":true,"supports_prompt_caching":true,"source":"https://docs.devin.ai/windsurf/plugins/cascade/models","provider":"cognition","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cognition/swe-1.7":{"mode":"chat","base_model":"swe-1.7","supports_function_calling":true,"supports_prompt_caching":true,"source":"https://docs.devin.ai/desktop/models","provider":"cognition","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cognition/swe-1.7-lightning":{"mode":"chat","base_model":"swe-1.7-lightning","supports_function_calling":true,"supports_prompt_caching":true,"source":"https://docs.devin.ai/desktop/models","provider":"cognition","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/glm-4.5-air":{"mode":"chat","base_model":"glm-4.5-air","max_input_tokens":128000,"max_output_tokens":128000,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":true,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":true,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/qwen3-30b-a3b":{"mode":"chat","base_model":"qwen3-30b-a3b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":true,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/qwen3-coder-30b-a3b":{"mode":"chat","base_model":"qwen3-coder-30b-a3b","max_input_tokens":131072,"max_output_tokens":131072,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":false,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":163840,"max_output_tokens":163840,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":true,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":163840,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"pinstripes/ps/minimax-m2.7":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":1000192,"max_output_tokens":1000192,"supports_function_calling":true,"supports_assistant_prefill":true,"supports_reasoning":false,"source":"https://pinstripes.io/","provider":"pinstripes","max_tokens":1000192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"darkbloom/gemma-4-26b":{"mode":"chat","base_model":"gemma-4-26b","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.darkbloom.dev/","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"darkbloom","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"darkbloom/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":32768,"source":"https://www.darkbloom.dev/","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"darkbloom","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-0309-non-reasoning":{"mode":"chat","base_model":"grok-4.20-0309-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-multi-agent-0309":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supported_endpoints":["/v1/responses"],"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-build-0.1":{"mode":"chat","base_model":"grok-build-0.1","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://docs.x.ai/developers/models/grok-build-0.1","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"provider":"xai","max_tokens":256000,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-mythos-5":{"mode":"chat","base_model":"claude-mythos-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"provider_specific_entry":{"us":1.1},"provider":"anthropic","web_search_billing_unit":"per_request","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-mythos-5-1":{"mode":"chat","base_model":"claude-mythos-5-1","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://platform.claude.com/docs/en/about-claude/pricing","provider":"anthropic","web_search_billing_unit":"per_request","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-mythos-preview":{"mode":"chat","base_model":"claude-mythos","deprecation_date":"2026-06-09","deprecation_source":"https://platform.claude.com/docs/en/about-claude/model-deprecations","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/models/overview","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"provider_specific_entry":{"us":1.1},"provider":"anthropic","web_search_billing_unit":"per_request","supports_safeguards":true,"is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-robotics-er-2-streaming-preview":{"mode":"chat","base_model":"gemini-robotics-er-2-streaming","max_input_tokens":131072,"max_output_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_reasoning":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"gemini","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-small-2603":{"mode":"chat","base_model":"mistral-small","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-small-4-0-26-03","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/labs-leanstral-1-5":{"mode":"chat","base_model":"labs-leanstral-1-5","deprecation_date":"2026-09-30","max_input_tokens":262144,"max_output_tokens":131072,"source":"https://docs.mistral.ai/models/model-cards/leanstral-1-5","supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash":{"mode":"chat","base_model":"deepseek-flash","token_count_framing":23,"max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://fireworks.ai/models/fireworks/deepseek-v4p1-flash","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":393216,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_endpoints":["/v1/chat/completions"],"supports_service_tier":true,"supports_system_messages":true,"supports_native_streaming":true,"metadata":{"notes":"DeepSeek-V4.1-Flash on Fireworks serverless (552B MoE, 1M context, image input). Priority serving tier is 1.25x Standard across all token types.","calculation":"Standard: $0.22/M uncached input, $0.007/M cached input, $0.66/M output. Priority: $0.275/M, $0.00875/M, $0.825/M."},"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":1048576,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/deepseek-v4p1-flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","token_count_framing":23,"max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision","deprecation_date":"2026-09-25","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":1048576,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p2-fast":{"mode":"chat","base_model":"glm-5.2-fast","deprecation_date":"2026-09-25","deprecation_source":"https://docs.fireworks.ai/changelog.md","successor_model":"fireworks_ai/glm-5p3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p2-fast-us":{"mode":"chat","base_model":"glm-5.2-fast-us","deprecation_date":"2026-09-25","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/glm-5p3","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k3-fast":{"mode":"chat","base_model":"kimi-k3-fast","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/kimi-k3-us":{"mode":"chat","base_model":"kimi-k3-us","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/qwen3p8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":262144,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/muse-glimmer-30b":{"mode":"chat","base_model":"muse-glimmer-30b","deprecation_date":"2026-09-25","max_input_tokens":131072,"max_output_tokens":16384,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":16384,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/nemotron-lightning-3p5-30b-a3b":{"mode":"chat","base_model":"nemotron-lightning-3.5-30b-a3b","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"Nemotron Super VL","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":32768,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/nemotron-3-ultra-nvfp4":{"mode":"chat","base_model":"nemotron-3-ultra-nvfp4","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","successor_model":"Nemotron Super VL","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":32768,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/muse-glimmer-30b":{"mode":"chat","base_model":"muse-glimmer-30b","deprecation_date":"2026-09-25","max_input_tokens":131072,"max_output_tokens":16384,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":16384,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b":{"mode":"chat","base_model":"nemotron-lightning-3.5-30b-a3b","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"Nemotron Super VL","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":32768,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/nemotron-3-ultra-nvfp4":{"mode":"chat","base_model":"nemotron-3-ultra-nvfp4","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://api.fireworks.ai/v1/serverless/models","successor_model":"Nemotron Super VL","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":32768,"is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/qwen3p8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":262144,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"supports_mid_conversation_system_messages":false,"supports_reasoning_content_blocks":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/glm-5p2-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/glm-5.2-fast","deprecation_date":"2026-09-25","deprecation_source":"https://docs.fireworks.ai/changelog.md","successor_model":"fireworks_ai/accounts/fireworks/models/glm-5p3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/glm-5p2-fast-us":{"mode":"chat","base_model":"accounts/fireworks/routers/glm-5.2-fast-us","deprecation_date":"2026-09-25","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://docs.fireworks.ai/serverless/pricing","successor_model":"fireworks_ai/accounts/fireworks/models/glm-5p3","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/kimi-k3-fast":{"mode":"chat","base_model":"accounts/fireworks/routers/kimi-k3-fast","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","supports_prompt_caching":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/routers/kimi-k3-us":{"mode":"chat","base_model":"accounts/fireworks/routers/kimi-k3-us","max_input_tokens":1048576,"max_output_tokens":131072,"reasoning_effort_levels":["low","high","max"],"source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/moonshotai/kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/tencent/hy3":{"mode":"chat","base_model":"hy3","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/moonshotai/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision","max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/mindai/macaron-v1-venti":{"mode":"chat","base_model":"mindai/macaron-v1-venti","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/minimax/minimax-m3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"max_output_tokens":393216,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":393216,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/inclusionai/ling-3.0-flash-fast":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash-fast","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/inclusionai/ling-3.0-flash":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/mindai/macaron-v1-tall":{"mode":"chat","base_model":"mindai/macaron-v1-tall","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/stepfun/step-3.7-flash":{"mode":"chat","base_model":"stepfun/step-3.7-flash","max_input_tokens":262144,"max_output_tokens":256000,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/nvidia/nemotron-3-nano-30b-a3b":{"mode":"chat","base_model":"nemotron-3-nano-30b-a3b","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/baidu/cobuddy":{"mode":"chat","base_model":"cobuddy","max_input_tokens":131072,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/xiaomimimo/mimo-v2.5":{"mode":"chat","base_model":"mimo","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","max_input_tokens":1000000,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/xiaomimimo/mimo-v2.5-pro":{"mode":"chat","base_model":"mimo-v2.5-pro","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.6-27b":{"mode":"chat","base_model":"qwen3.6-27b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/moonshotai/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5.1":{"mode":"chat","base_model":"glm-5.1","max_input_tokens":204800,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/minimax/minimax-m2.7-highspeed":{"mode":"chat","base_model":"minimax-m2.7-highspeed","max_input_tokens":204800,"max_output_tokens":131072,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5v-turbo":{"mode":"chat","base_model":"glm-5v-turbo","max_input_tokens":204800,"max_output_tokens":131072,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","max_input_tokens":262144,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/google/gemma-4-31b-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":262144,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5-turbo":{"mode":"chat","base_model":"glm-5-turbo","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/minimax/minimax-m2.7":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":204800,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/minimax/minimax-m2.5-highspeed":{"mode":"chat","base_model":"minimax-m2.5-highspeed","max_input_tokens":204800,"max_output_tokens":131100,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131100,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.5-27b":{"mode":"chat","base_model":"qwen3.5-27b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.5-122b-a10b":{"mode":"chat","base_model":"qwen3.5-122b-a10b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.5-35b-a3b":{"mode":"chat","base_model":"qwen3.5-35b-a3b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.5-397b-a17b":{"mode":"chat","base_model":"qwen3.5-397b-a17b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/minimax/minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":204800,"max_output_tokens":131100,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131100,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-5":{"mode":"chat","base_model":"glm-5","max_input_tokens":202800,"max_output_tokens":131072,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-ocr-2":{"mode":"chat","base_model":"deepseek-ocr-2","max_input_tokens":8192,"max_output_tokens":8192,"source":"https://novita.ai/pricing","supports_vision":true,"provider":"novita","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/moonshotai/kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-4.7-h":{"mode":"chat","base_model":"glm-4.7-h","max_input_tokens":204800,"max_output_tokens":131072,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/zai-org/glm-4.7-flash":{"mode":"chat","base_model":"glm-4.7-flash","max_input_tokens":200000,"max_output_tokens":128000,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/qwen/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://novita.ai/pricing","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"novita","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek_v3":{"mode":"chat","base_model":"deepseek-v3","max_input_tokens":64000,"max_output_tokens":16000,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-r1":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":64000,"max_output_tokens":16000,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-v3/community":{"mode":"chat","base_model":"deepseek-v3/community","max_input_tokens":64000,"max_output_tokens":8000,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/deepseek/deepseek-r1/community":{"mode":"chat","base_model":"deepseek-r1/community","max_input_tokens":64000,"max_output_tokens":8000,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/thudm/glm-4-32b-0414":{"mode":"chat","base_model":"thudm/glm-4-32b","max_input_tokens":32000,"max_output_tokens":32000,"source":"https://api.novita.ai/v3/openai/models","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"novita/meta-llama/llama-3.2-1b-instruct":{"mode":"chat","base_model":"llama-3.2-1b-instruct","max_input_tokens":131000,"max_output_tokens":32000,"source":"https://api.novita.ai/v3/openai/models","supports_response_schema":true,"supports_vision":false,"provider":"novita","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"deepseek-v4-flash","supports_reasoning":true,"max_output_tokens":1048576,"max_input_tokens":1049000,"deprecation_date":"2026-10-05","supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","supports_reasoning":true,"max_output_tokens":1048576,"max_input_tokens":1049000,"deprecation_date":"2026-10-05","supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31b-it","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/ibm-granite/granite-4.1-8b":{"mode":"chat","base_model":"granite-4.1-8b","deprecation_date":"2026-10-05","max_output_tokens":131072,"max_input_tokens":131000,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/JetBrains/Mellum2-12B-A2.5B-Instruct":{"mode":"chat","base_model":"jetbrains/mellum2-12b-a2.5b-instruct","deprecation_date":"2026-10-05","max_output_tokens":131072,"max_input_tokens":131000,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/meta-llama/Llama-3.1-70B-Instruct":{"mode":"chat","base_model":"llama-3.1-70b-instruct","deprecation_date":"2026-10-05","max_output_tokens":128000,"max_input_tokens":131000,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"minimax-m3","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"kimi-k2.7-code","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"kimi-k2.6","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B":{"mode":"chat","base_model":"nvidia-nemotron-3.5-lightning-30b-a3b","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"mode":"chat","base_model":"nvidia-nemotron-3-ultra-550b-a55b","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/OpenPipe/Qwen3-14B-Instruct":{"mode":"chat","base_model":"openpipe/qwen3-14b-instruct","deprecation_date":"2026-10-05","max_output_tokens":32768,"max_input_tokens":32800,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/Qwen/Qwen3.8-27B":{"mode":"chat","base_model":"qwen3.8-27b","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_prompt_caching":true,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/Qwen/Qwen3.6-35B-A3B":{"mode":"chat","base_model":"qwen3.6-35b-a3b","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","supports_prompt_caching":true,"provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/Qwen/Qwen3.6-27B":{"mode":"chat","base_model":"qwen3.6-27b","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"deprecation_date":"2026-10-05","supports_prompt_caching":true,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/Qwen/Qwen3.5-35B-A3B":{"mode":"chat","base_model":"qwen3.5-35b-a3b","deprecation_date":"2026-10-05","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":262000,"supports_vision":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/Qwen/Qwen3-30B-A3B-Instruct-2507":{"mode":"chat","base_model":"qwen3-30b-a3b-instruct","deprecation_date":"2026-10-05","max_output_tokens":262144,"max_input_tokens":262000,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","supports_reasoning":true,"max_input_tokens":1049000,"supports_prompt_caching":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/ibm-granite/granite-4.2-8b":{"mode":"chat","base_model":"granite-4.2-8b","supports_reasoning":true,"max_input_tokens":131000,"supports_prompt_caching":true,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","supports_reasoning":true,"max_output_tokens":262144,"max_input_tokens":1049000,"supports_prompt_caching":true,"supports_vision":false,"source":"https://wandb.ai/site/pricing/tokens/","provider":"wandb","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/openai/gpt-oss-120b-Turbo":{"mode":"chat","base_model":"gpt-oss-120b-turbo","max_output_tokens":131072,"max_input_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/MiniMaxAI/MiniMax-M2.7":{"mode":"chat","base_model":"minimax-m2.7","max_output_tokens":196608,"max_input_tokens":196608,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":196608,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.8-27B":{"mode":"chat","base_model":"qwen3.8-27b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemma-4-31B-it-Ultra":{"mode":"chat","base_model":"gemma-4-31b-it-ultra","max_output_tokens":131072,"max_input_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"kimi-k2.5","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-4.7-Flash":{"mode":"chat","base_model":"glm-4.7-flash","max_output_tokens":202752,"max_input_tokens":202752,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-4.6":{"mode":"chat","base_model":"glm-4.6","max_output_tokens":202752,"max_input_tokens":202752,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","max_input_tokens":1000000,"max_output_tokens":1000000,"prompt_cache_min_tokens":1024,"source":"https://deepinfra.com/pricing","supports_adaptive_thinking":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","max_output_tokens":1000000,"max_input_tokens":1000000,"prompt_cache_min_tokens":1024,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_adaptive_thinking":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","max_output_tokens":1000000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/XiaomiMiMo/MiMo-V2.5":{"mode":"chat","base_model":"mimo-v2.5","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3-Max":{"mode":"chat","base_model":"qwen3-max","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemma-4-31B-it-turbo":{"mode":"chat","base_model":"gemma-4-31b-it-turbo","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/thinkingmachines/Inkling-Small":{"mode":"chat","base_model":"thinkingmachines/inkling-small","max_output_tokens":524288,"max_input_tokens":524288,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/meta-models/Muse-Glimmer-30B":{"mode":"chat","base_model":"models/muse-glimmer-30b","max_output_tokens":131072,"max_input_tokens":131072,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3-Max-Thinking":{"mode":"chat","base_model":"qwen3-max-thinking","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-instruct","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.5-27B":{"mode":"chat","base_model":"qwen3.5-27b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.6-35B-A3B":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/nvidia/Nemotron-Content-Safety-3.5":{"mode":"chat","base_model":"nemotron-content-safety-3.5","max_output_tokens":131072,"max_input_tokens":131072,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","max_input_tokens":1000000,"max_output_tokens":1000000,"prompt_cache_min_tokens":512,"source":"https://deepinfra.com/pricing","supports_adaptive_thinking":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/thinkingmachines/Inkling":{"mode":"chat","base_model":"thinkingmachines/inkling","max_output_tokens":524288,"max_input_tokens":524288,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"kimi-k2.6","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.7-Max":{"mode":"chat","base_model":"qwen3.7-max","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/ByteDance/Seed-2.0-mini":{"mode":"chat","base_model":"seed-2.0-mini","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.8-2.4T-A95B":{"mode":"chat","base_model":"qwen3.8-2.4t-a95b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"minimax-m3","max_output_tokens":524288,"max_input_tokens":524288,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemini-3.1-flash-lite":{"mode":"chat","base_model":"gemini-3.1-flash-lite","max_output_tokens":1000000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","max_output_tokens":1000000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_vision":true,"source":"https://deepinfra.com/pricing","supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/inclusionAI/Ling-3.0-flash":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash","max_output_tokens":131072,"max_input_tokens":131072,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/stepfun-ai/Step-3.7-Flash":{"mode":"chat","base_model":"stepfun-ai/step-3.7-flash","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.5-35B-A3B":{"mode":"chat","base_model":"qwen3.5-35b-a3b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/ByteDance/Seed-1.8":{"mode":"chat","base_model":"seed-1.8","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/tencent/Hy3":{"mode":"chat","base_model":"hy3","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/ByteDance/Seed-2.0-code":{"mode":"chat","base_model":"seed-2.0-code","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/ByteDance/Seed-2.0-pro":{"mode":"chat","base_model":"seed-2.0-pro","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-5":{"mode":"chat","base_model":"glm-5","max_output_tokens":202752,"max_input_tokens":202752,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/nvidia/Nemotron-3-Nano-30B-A3B":{"mode":"chat","base_model":"nemotron-3-nano-30b-a3b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"kimi-k2.7-code","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":1000000,"prompt_cache_min_tokens":1024,"source":"https://deepinfra.com/pricing","supports_adaptive_thinking":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"qwen3.5-397b-a17b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemma-4-E4B-it":{"mode":"chat","base_model":"gemma-4-e4b-it","max_output_tokens":131072,"max_input_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V3.2":{"mode":"chat","base_model":"deepseek-v3.2","max_output_tokens":163840,"max_input_tokens":163840,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":163840,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.8-Max":{"mode":"chat","base_model":"qwen3.8-max","max_output_tokens":256000,"max_input_tokens":256000,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":1000000,"prompt_cache_min_tokens":512,"source":"https://deepinfra.com/pricing","supports_adaptive_thinking":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"mode":"chat","base_model":"nvidia-nemotron-3-ultra-550b-a55b","max_output_tokens":262144,"max_input_tokens":262144,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.5-122B-A10B":{"mode":"chat","base_model":"qwen3.5-122b-a10b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-5.1":{"mode":"chat","base_model":"glm-5.1","max_output_tokens":202752,"max_input_tokens":202752,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B":{"mode":"chat","base_model":"nvidia-nemotron-3-super-120b-a12b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/moonshotai/Kimi-K3":{"mode":"chat","base_model":"kimi-k3","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","max_input_tokens":1000000,"max_output_tokens":1000000,"prompt_cache_min_tokens":2048,"source":"https://deepinfra.com/pricing","supports_adaptive_thinking":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.6-27B":{"mode":"chat","base_model":"qwen3.6-27b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemma-4-26B-A4B-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemini-3.1-pro":{"mode":"chat","base_model":"gemini-3.1-pro","max_output_tokens":1000000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","supports_audio_input":true,"provider":"deepinfra","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/XiaomiMiMo/MiMo-V2.5-Pro":{"mode":"chat","base_model":"mimo-v2.5-pro","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/anthropic/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_output_tokens":200000,"max_input_tokens":200000,"prompt_cache_min_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":200000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/openai/gpt-oss-120b-Ultra":{"mode":"chat","base_model":"gpt-oss-120b-ultra","max_output_tokens":131072,"max_input_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/Qwen/Qwen3.5-9B":{"mode":"chat","base_model":"qwen3.5-9b","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/MiniMaxAI/MiniMax-M2.7-Turbo":{"mode":"chat","base_model":"minimax-m2.7-turbo","max_output_tokens":196608,"max_input_tokens":196608,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":196608,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/zai-org/GLM-4.7":{"mode":"chat","base_model":"glm-4.7","max_output_tokens":202752,"max_input_tokens":202752,"supports_prompt_caching":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":202752,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepinfra/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/pricing","provider":"deepinfra","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-omni-1.1-flash":{"mode":"chat","base_model":"gemini-omni-1.1-flash","deprecation_date":"2026-12-03","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","max_input_tokens":1048576,"max_output_tokens":65536,"rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/pricing","successor_model":"gemini/gemini-omni-1.2-flash","supported_endpoints":["/v1beta/interactions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","video"],"supports_audio_input":true,"supports_reasoning":true,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"tpm":800000,"provider":"gemini","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-reasoning-latest":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-non-reasoning":{"mode":"chat","base_model":"grok-4.20-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-non-reasoning-latest":{"mode":"chat","base_model":"grok-4.20-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-multi-agent":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supported_endpoints":["/v1/responses"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-4.20-multi-agent-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supported_endpoints":["/v1/responses"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"groq/qwen/qwen3.8-27b":{"mode":"chat","base_model":"qwen3.8-27b","max_input_tokens":131042,"max_output_tokens":16384,"source":"https://console.groq.com/docs/model/qwen/qwen3.8-27b","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"groq","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-medium-3.5":{"mode":"chat","base_model":"mistral-medium-3.5","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-vibe-cli-latest":{"mode":"chat","base_model":"mistral-vibe-cli","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-vibe-cli-with-tools":{"mode":"chat","base_model":"mistral-vibe-cli-with-tools","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-vibe-cli-fast":{"mode":"chat","base_model":"mistral-vibe-cli-fast","max_input_tokens":262144,"max_output_tokens":262144,"reasoning_effort_levels":["none","high"],"source":"https://docs.mistral.ai/models/model-cards/mistral-small-4-0-26-03","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-code-latest":{"mode":"chat","base_model":"mistral-code","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.mistral.ai/models/model-cards/codestral-25-08","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-code-fim-latest":{"mode":"chat","base_model":"mistral-code-fim","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://docs.mistral.ai/models/model-cards/codestral-25-08","supports_assistant_prefill":true,"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/mistral-code-agent-latest":{"mode":"chat","base_model":"mistral-code-agent","max_input_tokens":256000,"max_output_tokens":256000,"reasoning_effort_levels":["none","high"],"source":"https://mistral.ai/news/devstral-2-vibe-cli","supports_assistant_prefill":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"mistral","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"mistral/labs-leanstral-1-5-1":{"mode":"chat","base_model":"labs-leanstral-1-5-1","max_input_tokens":262144,"max_output_tokens":131072,"source":"https://docs.mistral.ai/models/model-cards/leanstral-1-5","supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":128000,"max_tokens":128000,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","supports_prompt_caching":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/glm-5p3-flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"fireworks_ai/accounts/fireworks/models/inkling":{"mode":"chat","base_model":"inkling","deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://fireworks.ai/models/fireworks/inkling","successor_model":"fireworks_ai/accounts/fireworks/models/muse-spark-open","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","is_deprecated":true,"max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"zai/glm-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://docs.z.ai/guides/overview/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"zai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen3.8-Flash":{"mode":"chat","base_model":"qwen3.8-flash","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.together.ai/v1/models","provider":"together_ai","max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/moonshotai/Kimi-K2.5-fp4":{"mode":"chat","base_model":"kimi-k2.5-fp4","max_input_tokens":262144,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/MiniMaxAI/MiniMax-M2.7":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":196608,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"deepseek-r1","max_input_tokens":163840,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/mistralai/Ministral-3-14B-Instruct-2512":{"mode":"chat","base_model":"ministral-3-14b-instruct","max_input_tokens":262144,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/nvidia/NVIDIA-Nemotron-Nano-9B-v2":{"mode":"chat","base_model":"nvidia-nemotron-nano-9b-v2","deprecation_date":"2026-02-25","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":131072,"source":"https://api.together.ai/v1/models","provider":"together_ai","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/mistralai/Mistral-7B-Instruct-v0.3":{"mode":"chat","base_model":"mistral-7b-instruct","max_input_tokens":32768,"source":"https://api.together.ai/v1/models","provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cerebras/gemma-4-31b":{"mode":"chat","base_model":"gemma-4-31b","max_input_tokens":131072,"max_output_tokens":40960,"source":"https://api.cerebras.ai/public/v1/models/gemma-4-31b","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"cerebras","max_tokens":40960,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/glm-5.2":{"mode":"chat","base_model":"glm-5.2","deprecation_date":"2026-11-16","max_input_tokens":256000,"max_output_tokens":16384,"source":"https://www.scaleway.com/en/pricing/model-as-a-service/","supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"provider":"scaleway","is_deprecated":true,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"scaleway/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":256000,"max_output_tokens":32768,"source":"https://www.scaleway.com/en/pricing/model-as-a-service/","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false,"provider":"scaleway","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","deprecation_date":"2026-10-03","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":262144,"max_output_tokens":262144,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]}]},"bedrock/us-gov-west-1/nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nemotron-nano-3-30b","max_input_tokens":256000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nemotron-nano-12b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_system_messages":true,"supports_vision":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nemotron-nano-9b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_system_messages":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":16384,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":16384,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-west-1/anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"token_count_model":{"model":"anthropic.claude-fable-5-1","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nemotron-nano-3-30b","max_input_tokens":256000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nemotron-nano-12b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supports_system_messages":true,"supports_vision":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nemotron-nano-9b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supports_system_messages":true,"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"provider":"bedrock","max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":16384,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":128000,"max_output_tokens":16384,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_inline_reasoning":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock","max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us-gov-east-1/anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-5.6-terra":{"mode":"responses","base_model":"gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-5.6-luna":{"mode":"responses","base_model":"gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-5.4":{"mode":"responses","base_model":"gpt-5.4","max_input_tokens":272000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1"],"max_tokens":128000,"supports_service_tier":true,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/xai.grok-4.3":{"mode":"responses","base_model":"grok-4.3","use_openai_responses_path":true,"max_input_tokens":131072,"max_output_tokens":16384,"supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-east-1","us-east-2","us-gov-west-1"],"max_tokens":16384,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","use_openai_responses_path":true,"max_input_tokens":500000,"max_output_tokens":500000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-gov-east-1","us-east-1","us-east-2","us-west-1","us-gov-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":500000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/google.gemma-4-e2b":{"mode":"chat","base_model":"gemma-4-e2b","max_input_tokens":128000,"max_output_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":true,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/google.gemma-4-26b-a4b":{"mode":"chat","base_model":"gemma-4-26b-a4b","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/google.gemma-4-31b":{"mode":"chat","base_model":"gemma-4-31b","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aws.amazon.com/bedrock/pricing/","use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1"],"max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-5.4":{"mode":"responses","base_model":"gpt-5.4","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1"],"max_tokens":128000,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-east-1/xai.grok-4.6":{"mode":"chat","base_model":"grok-4.6","use_openai_responses_path":true,"max_input_tokens":500000,"max_output_tokens":500000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-west-2","us-gov-east-1","us-east-1","us-east-2","us-west-1","us-gov-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":500000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","supports_vision":false,"max_input_tokens":131072,"max_output_tokens":32768,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/us-gov/gpt-5.1":{"mode":"chat","base_model":"us-gov/gpt-5.1","default_reasoning_effort":"none","max_input_tokens":272000,"max_output_tokens":128000,"supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","max_tokens":128000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us-gov/o3-mini":{"mode":"chat","base_model":"us-gov/o3-mini","max_input_tokens":200000,"max_output_tokens":100000,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","max_tokens":100000,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions"]}]},"gemini/lyria-3.5-clip-preview":{"mode":"chat","base_model":"lyria-3.5-clip","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/lyria-3.5-pro-preview":{"mode":"chat","base_model":"lyria-3.5-pro","max_input_tokens":131072,"max_output_tokens":8192,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"provider":"gemini","max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/lyria-3.5":{"mode":"chat","base_model":"lyria-3.5","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"provider":"gemini","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"perplexity/anthropic/claude-fable-5":{"mode":"responses","base_model":"claude-fable-5","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-opus-5":{"mode":"responses","base_model":"claude-opus-5","supports_adaptive_thinking":true,"prompt_cache_min_tokens":512,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-opus-4-8":{"mode":"responses","base_model":"claude-opus-4-8","supports_adaptive_thinking":true,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-sonnet-5":{"mode":"responses","base_model":"claude-sonnet-5","supports_adaptive_thinking":true,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/anthropic/claude-sonnet-4-6":{"mode":"responses","base_model":"claude-sonnet-4-6","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.6-sol":{"mode":"responses","base_model":"gpt-5.6-sol","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.6-terra":{"mode":"responses","base_model":"gpt-5.6-terra","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.6-luna":{"mode":"responses","base_model":"gpt-5.6-luna","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.5":{"mode":"responses","base_model":"gpt-5.5","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.4":{"mode":"responses","base_model":"gpt-5.4","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.4-mini":{"mode":"responses","base_model":"gpt-5.4-mini","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5.4-nano":{"mode":"responses","base_model":"gpt-5.4-nano","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/openai/gpt-5":{"mode":"responses","base_model":"gpt-5","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.1-pro-preview":{"mode":"responses","base_model":"gemini-3.1-pro","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.1-flash-lite":{"mode":"responses","base_model":"gemini-3.1-flash-lite","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.5-flash":{"mode":"responses","base_model":"gemini-3.5-flash","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.5-flash-lite":{"mode":"responses","base_model":"gemini-3.5-flash-lite","supports_sampling_params":false,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.6-flash":{"mode":"responses","base_model":"gemini-3.6-flash","supports_sampling_params":false,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/google/gemini-3.7-flash":{"mode":"responses","base_model":"gemini-3.7-flash","supports_sampling_params":false,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.6":{"mode":"responses","base_model":"grok-4.6","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.5":{"mode":"responses","base_model":"grok-4.5","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.3":{"mode":"responses","base_model":"grok-4.3","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.20-reasoning":{"mode":"responses","base_model":"grok-4.20","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.20-non-reasoning":{"mode":"responses","base_model":"grok-4.20-non","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/xai/grok-4.20-multi-agent":{"mode":"responses","base_model":"grok-4.20-multi-agent","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/glm-5.3":{"mode":"responses","base_model":"glm-5.3","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/glm-5.3-flash":{"mode":"responses","base_model":"glm-5.3-flash","supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/nemotron-3.5-lightning-30b-a3b":{"mode":"responses","base_model":"nemotron-3.5-lightning-30b-a3b","supports_web_search":true,"supports_function_calling":true,"supports_reasoning":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"perplexity/perplexity/nemotron-3-ultra-550b-a55b":{"mode":"responses","base_model":"nemotron-3-ultra-550b-a55b","supports_web_search":true,"supports_function_calling":true,"supports_reasoning":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/responses"]}]},"openrouter/anthropic/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_sampling_params":false,"supports_adaptive_thinking":true,"thinking_always_on":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"prompt_cache_min_tokens":512,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-fable-5.1":{"mode":"chat","base_model":"claude-fable-5.1","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_sampling_params":false,"supports_adaptive_thinking":true,"thinking_always_on":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"prompt_cache_min_tokens":512,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.8":{"mode":"chat","base_model":"claude-opus-4.8","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_sampling_params":false,"supports_adaptive_thinking":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_sampling_params":false,"supports_adaptive_thinking":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supports_code_execution":true,"supports_parallel_function_calling":true,"supports_service_tier":true,"supports_system_messages":true,"supports_web_search":true,"source":"https://openrouter.ai/api/v1/models","supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_service_tier":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_service_tier":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_service_tier":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-luna-pro":{"mode":"chat","base_model":"gpt-5.6-luna-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-terra-pro":{"mode":"chat","base_model":"gpt-5.6-terra-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.20":{"mode":"chat","base_model":"grok-4.20","provider":"openrouter","max_input_tokens":2000000,"max_output_tokens":1800000,"max_tokens":1800000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1800000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.20-multi-agent":{"mode":"chat","base_model":"grok-4.20-multi-agent","provider":"openrouter","max_input_tokens":2000000,"max_output_tokens":1800000,"max_tokens":1800000,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_function_calling":false,"supports_tool_choice":false,"supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1800000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.3":{"mode":"chat","base_model":"grok-4.3","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":900000,"max_tokens":900000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":900000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.5":{"mode":"chat","base_model":"grok-4.5","provider":"openrouter","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":450000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.6":{"mode":"chat","base_model":"grok-4.6","provider":"openrouter","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":450000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-build-0.1":{"mode":"chat","base_model":"grok-build-0.1","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":230400,"max_tokens":230400,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":230400},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m3":{"mode":"chat","base_model":"minimax-m3","provider":"openrouter","max_input_tokens":524288,"max_output_tokens":512000,"max_tokens":512000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/openai/gpt-6-astra","supports_audio_input":false,"supports_web_search":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-astra-pro":{"mode":"chat","base_model":"gpt-6-astra-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/openai/gpt-6-astra-pro","supports_audio_input":false,"supports_web_search":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/qwen/qwen3.8-flash":{"mode":"chat","base_model":"qwen3.8-flash","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.3-flash":{"mode":"chat","base_model":"glm-5.3-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision-exp","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.3":{"mode":"chat","base_model":"glm-5.3","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-27b":{"mode":"chat","base_model":"qwen3.8-27b","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-2.4t-a95b":{"mode":"chat","base_model":"qwen3.8-2.4t-a95b","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3.5-lightning:free":{"mode":"chat","base_model":"nemotron-3.5-lightning","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"source":"https://openrouter.ai/qwen/qwen3.8-max","supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-max-0902":{"mode":"chat","base_model":"qwen3.8-max-0902","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash-0731","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.7-flash":{"mode":"chat","base_model":"qwen3.7-flash","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/poolside/laguna-s-2.1":{"mode":"chat","base_model":"laguna-s-2.1","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"supports_prompt_caching":true,"supports_response_schema":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/poolside/laguna-s-2.1:free":{"mode":"chat","base_model":"poolside/laguna-s-2.1","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k3":{"mode":"chat","base_model":"kimi-k3","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/poolside/laguna-xs-2.1":{"mode":"chat","base_model":"laguna-xs-2.1","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"supports_prompt_caching":true,"supports_response_schema":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/poolside/laguna-xs-2.1:free":{"mode":"chat","base_model":"poolside/laguna-xs-2.1","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-lite-image":{"mode":"chat","base_model":"gemini-3.1-flash-lite-image","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":58982},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-image":{"mode":"chat","base_model":"gemini-3.1-flash-image","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3-pro-image":{"mode":"chat","base_model":"gemini-3-pro-image","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.2":{"mode":"chat","base_model":"glm-5.2","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.2:free":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":29491},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3.5-content-safety":{"mode":"chat","base_model":"nemotron-3.5-content-safety","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_reasoning":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_response_schema":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3.5-content-safety:free":{"mode":"chat","base_model":"nemotron-3.5-content-safety","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_response_schema":false,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-ultra-550b-a55b":{"mode":"chat","base_model":"nemotron-3-ultra-550b-a55b","provider":"openrouter","max_input_tokens":202800,"max_output_tokens":182520,"max_tokens":182520,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":182520},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-ultra-550b-a55b:free":{"mode":"chat","base_model":"nemotron-3-ultra-550b-a55b","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m3:free":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"source":"https://openrouter.ai/minimax/minimax-m3:free","supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-medium-3-5":{"mode":"chat","base_model":"mistral-medium-3-5","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free":{"mode":"chat","base_model":"nemotron-3-nano-omni-30b-a3b","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_audio_input":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-plus-20260420":{"mode":"chat","base_model":"qwen3.5-plus-20260420","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.6-flash":{"mode":"chat","base_model":"qwen3.6-flash","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.6-max-preview":{"mode":"chat","base_model":"qwen3.6-max-preview","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.6-27b":{"mode":"chat","base_model":"qwen3.6-27b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262140,"max_tokens":262140,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262140},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.5-pro":{"mode":"chat","base_model":"gpt-5.5-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-chat-latest":{"mode":"chat","base_model":"gpt-chat-latest","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","provider":"openrouter","max_input_tokens":1024000,"max_output_tokens":384000,"max_tokens":384000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/moonshotai/kimi-k2.5","supports_video_input":true,"supports_audio_input":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemma-4-26b-a4b-it:free":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemma-4-31b-it":{"mode":"chat","base_model":"gemma-4-31b-it","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemma-4-31b-it:free":{"mode":"chat","base_model":"gemma-4-31b-it","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5v-turbo":{"mode":"chat","base_model":"glm-5v-turbo","provider":"openrouter","max_input_tokens":202752,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"deprecation_date":"2098-12-31","source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m2.7":{"mode":"chat","base_model":"minimax-m2.7","provider":"openrouter","max_input_tokens":204800,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m2.7:free":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":196608,"max_output_tokens":176947,"max_tokens":176947,"source":"https://openrouter.ai/minimax/minimax-m2.7:free","supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":176947},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-small-2603":{"mode":"chat","base_model":"mistral-small-2603","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5-turbo":{"mode":"chat","base_model":"glm-5-turbo","provider":"openrouter","max_input_tokens":202752,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"deprecation_date":"2098-12-31","source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-super-120b-a12b":{"mode":"chat","base_model":"nemotron-3-super-120b-a12b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-super-120b-a12b:free":{"mode":"chat","base_model":"nemotron-3-super-120b-a12b","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.5-9b":{"mode":"chat","base_model":"qwen3.5-9b","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-pro":{"mode":"chat","base_model":"gpt-5.4-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_service_tier":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-image-preview":{"mode":"chat","base_model":"gemini-3.1-flash-image-preview","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":58982},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-pro-preview-customtools":{"mode":"chat","base_model":"gemini-3.1-pro-preview-customtools","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-max-thinking":{"mode":"chat","base_model":"qwen3-max-thinking","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/minimax-m2-her":{"mode":"chat","base_model":"minimax-m2-her","max_input_tokens":65536,"max_output_tokens":2048,"max_tokens":2048,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_tool_choice":false,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":2048},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-audio":{"mode":"chat","base_model":"gpt-audio","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_audio_input":true,"supports_audio_output":true,"source":"https://openrouter.ai/api/v1/models","supports_vision":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-audio-mini":{"mode":"chat","base_model":"gpt-audio-mini","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_audio_input":true,"supports_audio_output":true,"source":"https://openrouter.ai/api/v1/models","supports_vision":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3-nano-30b-a3b":{"mode":"chat","base_model":"nemotron-3-nano-30b-a3b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-4.6v":{"mode":"chat","base_model":"glm-4.6v","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3-pro-image-preview":{"mode":"chat","base_model":"gemini-3-pro-image-preview","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":32768,"max_tokens":32768,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.1-codex":{"mode":"chat","base_model":"gpt-5.1-codex","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.1-codex-mini":{"mode":"chat","base_model":"gpt-5.1-codex-mini","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k2-thinking":{"mode":"chat","base_model":"kimi-k2-thinking","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":98304,"max_tokens":98304,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":98304},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/voxtral-small-24b-2507":{"mode":"chat","base_model":"voxtral-small-24b-2507","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":26214,"max_tokens":26214,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_vision":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":26214},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-safeguard-20b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-vl-32b-instruct":{"mode":"chat","base_model":"qwen3-vl-32b-instruct","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-vl-8b-thinking":{"mode":"chat","base_model":"qwen3-vl-8b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-vl-8b-instruct":{"mode":"chat","base_model":"qwen3-vl-8b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-2.5-flash-image":{"mode":"chat","base_model":"gemini-2.5-flash-image","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":8192,"max_tokens":8192,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"deprecation_date":"2027-03-15","source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_tool_choice":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-vl-30b-a3b-thinking":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"openrouter/qwen/qwen3-vl-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-vl-30b-a3b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-pro":{"mode":"chat","base_model":"gpt-5-pro","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-vl-235b-a22b-thinking":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-thinking","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true},"openrouter/qwen/qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-max":{"mode":"chat","base_model":"qwen3-max","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"openrouter/deepseek/deepseek-v3.1-terminus":{"mode":"chat","base_model":"deepseek-v3.1-terminus","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"is_deprecated":true,"deprecation_date":"2026-09-28","source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_prompt_caching":true,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-coder-flash":{"mode":"chat","base_model":"qwen3-coder-flash","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-next-80b-a3b-thinking":{"mode":"chat","base_model":"qwen3-next-80b-a3b-thinking","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"openrouter/qwen/qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_prompt_caching":false,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen-plus-2025-07-28":{"mode":"chat","base_model":"qwen-plus-2025-07-28","max_input_tokens":1000000,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k2-0905":{"mode":"chat","base_model":"kimi-k2-0905","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":98304,"max_tokens":98304,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":98304},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-30b-a3b-thinking-2507":{"mode":"chat","base_model":"qwen3-30b-a3b-thinking-2507","max_input_tokens":81920,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-medium-3.1":{"mode":"chat","base_model":"mistral-medium-3.1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":104857,"max_tokens":104857,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":104857},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-4.5v":{"mode":"chat","base_model":"glm-4.5v","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-coder-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/arcee-ai/trinity-mini":{"mode":"chat","base_model":"arcee-ai/trinity-mini","source":"https://api.together.ai/v1/models","max_input_tokens":128000,"max_output_tokens":128000,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-omni-1.1-flash":{"mode":"chat","base_model":"gemini-omni-1.1-flash","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","successor_model":"vertex_ai/gemini-omni-1.2-flash","provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-omni-1.1-flash-preview":{"mode":"chat","base_model":"gemini-omni-1.1-flash","max_input_tokens":131072,"max_output_tokens":57920,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","successor_model":"vertex_ai/gemini-omni-1.2-flash","supported_endpoints":["/v1beta/interactions"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","video"],"supports_reasoning":true,"supports_video_input":true,"supports_vision":true,"provider":"vertex_ai","max_tokens":57920,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":57920},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4","source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","provider":"vertex_ai","max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":250000,"rpm":10,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gpt-5.5-cyber":{"mode":"chat","base_model":"gpt-5.5-cyber","fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_reasoning":true,"provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"gpt-rosalind-research":{"mode":"chat","base_model":"gpt-rosalind-research","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"together_ai/meta-llama/Llama-3.1-405B-Instruct":{"mode":"chat","base_model":"llama-3.1-405b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":4096,"max_output_tokens":4096,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/meta-llama/Llama-3.2-1B-Instruct":{"mode":"chat","base_model":"llama-3.2-1b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":131072,"max_output_tokens":131072,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/meta-llama/Llama-3.2-3B-Instruct":{"mode":"chat","base_model":"llama-3.2-3b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":131072,"max_output_tokens":131072,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen2-1.5B-Instruct":{"mode":"chat","base_model":"qwen2-1.5b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":32768,"max_output_tokens":32768,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen2.5-14B-Instruct":{"mode":"chat","base_model":"qwen2.5-14b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":32768,"max_output_tokens":32768,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"together_ai/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"qwen2.5-72b-instruct","source":"https://api.together.ai/v1/models","max_input_tokens":32768,"max_output_tokens":32768,"provider":"together_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/eu/codex-mini":{"mode":"chat","base_model":"codex-mini","deprecation_date":"2026-11-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","is_deprecated":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/computer-use-preview":{"mode":"chat","base_model":"computer-use","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-nano","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-10-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","is_deprecated":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5-codex":{"mode":"chat","base_model":"gpt-5-codex","deprecation_date":"2027-03-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1-codex","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4-nano","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5-pro":{"mode":"chat","base_model":"gpt-5-pro","deprecation_date":"2027-04-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.2-pro","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.1-codex-max":{"mode":"chat","base_model":"gpt-5.1-codex-max","deprecation_date":"2027-05-18","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","deprecation_date":"2027-06-08","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.4","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.2-codex":{"mode":"chat","base_model":"gpt-5.2-codex","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.3-codex","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.2-pro":{"mode":"chat","base_model":"gpt-5.2-pro","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","default_reasoning_effort":"none","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","default_reasoning_effort":"none","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-5.4-pro":{"mode":"chat","base_model":"gpt-5.4-pro","deprecation_date":"2027-09-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","deprecation_date":"2028-01-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/o1-mini":{"mode":"chat","base_model":"o1-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/o3-2025-04-16":{"mode":"chat","base_model":"o3","deprecation_date":"2026-11-19","successor_model":"azure/eu/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/o3-deep-research":{"mode":"chat","base_model":"o3","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6-sol","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/eu/o4-mini-2025-04-16":{"mode":"chat","base_model":"o4-mini","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/codex-mini":{"mode":"chat","base_model":"codex-mini","deprecation_date":"2026-11-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","is_deprecated":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/computer-use-preview":{"mode":"chat","base_model":"computer-use","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-4.1":{"mode":"chat","base_model":"gpt-4.1","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-4.1-mini":{"mode":"chat","base_model":"gpt-4.1-mini","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-4.1-nano":{"mode":"chat","base_model":"gpt-4.1-nano","deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-nano","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-4o-2024-05-13":{"mode":"chat","base_model":"gpt-4o","deprecation_date":"2026-10-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":128000,"max_output_tokens":4096,"source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","is_deprecated":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5":{"mode":"chat","base_model":"gpt-5","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5-codex":{"mode":"chat","base_model":"gpt-5-codex","deprecation_date":"2027-03-17","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1-codex","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5-mini":{"mode":"chat","base_model":"gpt-5-mini","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","deprecation_date":"2027-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4-nano","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5-pro":{"mode":"chat","base_model":"gpt-5-pro","deprecation_date":"2027-04-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.2-pro","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.1-codex-max":{"mode":"chat","base_model":"gpt-5.1-codex-max","deprecation_date":"2027-05-18","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","deprecation_date":"2027-06-08","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.4","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.2-codex":{"mode":"chat","base_model":"gpt-5.2-codex","deprecation_date":"2027-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.3-codex","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.2-pro":{"mode":"chat","base_model":"gpt-5.2-pro","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","default_reasoning_effort":"none","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","default_reasoning_effort":"none","deprecation_date":"2027-09-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/gpt-5.4-pro":{"mode":"chat","base_model":"gpt-5.4-pro","deprecation_date":"2027-09-07","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/o1-mini":{"mode":"chat","base_model":"o1-mini","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"azure/us/o3-deep-research":{"mode":"chat","base_model":"o3","deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6-sol","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions"]}]},"aihubmix/agnes-2.5-flash":{"mode":"chat","base_model":"agnes-2.5-flash","max_input_tokens":512000,"max_output_tokens":65500,"source":"https://aihubmix.com/api/v1/models","supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","max_tokens":65500,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65500},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/agnes-2.5-pro":{"mode":"chat","base_model":"agnes-2.5-pro","max_input_tokens":1000000,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/cc-glm-5.1":{"mode":"chat","base_model":"cc-glm-5.1","max_input_tokens":200000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_sampling_params":false,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":64000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"provider":"aihubmix","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/claude-opus-4-8-think":{"mode":"chat","base_model":"claude-opus-4-8-think","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"supports_sampling_params":false,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"supports_adaptive_thinking":true,"prompt_cache_min_tokens":512,"supports_sampling_params":false,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","max_input_tokens":1000000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"supports_adaptive_thinking":true,"supports_sampling_params":false,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/coding-glm-5.3":{"mode":"chat","base_model":"coding-glm-5.3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/coding-kimi-k3":{"mode":"chat","base_model":"coding-kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/coding-xiaomi-mimo-v2-omni":{"mode":"chat","base_model":"coding-xiaomi-mimo-v2-omni","source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/coding-xiaomi-mimo-v2.5":{"mode":"chat","base_model":"coding-xiaomi-mimo","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/coding-xiaomi-mimo-v2.5-pro":{"mode":"chat","base_model":"coding-xiaomi-mimo-v2.5-pro","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/command-a-plus-05-2026":{"mode":"chat","base_model":"command-a-plus","max_input_tokens":128000,"max_output_tokens":64000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":384000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":384000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/doubao-seed-2-0-code-preview":{"mode":"chat","base_model":"doubao-seed-2-0-code","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/doubao-seed-2-0-lite-260428":{"mode":"chat","base_model":"doubao-seed-2-0-lite","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/doubao-seed-2-0-mini":{"mode":"chat","base_model":"doubao-seed-2-0-mini","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/doubao-seed-2-0-pro":{"mode":"chat","base_model":"doubao-seed-2-0-pro","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/doubao-seed-2-1-turbo":{"mode":"chat","base_model":"doubao-seed-2-1-turbo","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/ernie-5.1":{"mode":"chat","base_model":"ernie-5.1","max_input_tokens":119000,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_reasoning":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3-flash-preview":{"mode":"chat","base_model":"gemini-3-flash","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3-flash-preview-search":{"mode":"chat","base_model":"gemini-3-flash-search","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3.1-pro-preview":{"mode":"chat","base_model":"gemini-3.1-pro","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3.1-pro-preview-customtools":{"mode":"chat","base_model":"gemini-3.1-pro-customtools","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","max_input_tokens":1048576,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"gemma-4-26b-a4b-it","max_input_tokens":262144,"max_output_tokens":131100,"source":"https://aihubmix.com/api/v1/models","supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","max_tokens":131100,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gemma-4-31b-it":{"mode":"chat","base_model":"gemma-4-31b-it","max_input_tokens":262144,"max_output_tokens":131100,"source":"https://aihubmix.com/api/v1/models","supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","max_tokens":131100,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/glm-5.2-fast-preview":{"mode":"chat","base_model":"glm-5.2-fast","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/glm-5.3":{"mode":"chat","base_model":"glm-5.3","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/glm-5.3-flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/glm-5v-turbo":{"mode":"chat","base_model":"glm-5v-turbo","max_input_tokens":200000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","max_input_tokens":400000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.4-high":{"mode":"chat","base_model":"gpt-5.4-high","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.4-low":{"mode":"chat","base_model":"gpt-5.4-low","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","max_input_tokens":400000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","max_input_tokens":400000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.5-pro":{"mode":"chat","base_model":"gpt-5.5-pro","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.6-sol-disc":{"mode":"chat","base_model":"gpt-5.6-sol-disc","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/gpt-chat-latest":{"mode":"chat","base_model":"gpt-chat","max_input_tokens":400000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/grok-4-20-non-reasoning":{"mode":"chat","base_model":"grok-4-20-non","max_input_tokens":1000000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/grok-4-20-reasoning":{"mode":"chat","base_model":"grok-4-20","max_input_tokens":1000000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/grok-4.6":{"mode":"chat","base_model":"grok-4.6","max_input_tokens":500000,"max_output_tokens":500000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":500000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":500000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/grok-build-0.1":{"mode":"chat","base_model":"grok-build-0.1","max_input_tokens":256000,"max_output_tokens":256000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":256000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/hy3":{"mode":"chat","base_model":"hy3","max_input_tokens":256000,"max_output_tokens":128000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/hy4-preview":{"mode":"chat","base_model":"hy4","max_input_tokens":1048576,"max_output_tokens":64000,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/kimi-k2.7-code-highspeed":{"mode":"chat","base_model":"kimi-k2.7-code-highspeed","max_input_tokens":262144,"max_output_tokens":32768,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":1048576,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/longcat-2.0":{"mode":"chat","base_model":"longcat-2.0","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/mai-thinking-1":{"mode":"chat","base_model":"mai-thinking-1","max_input_tokens":256000,"max_output_tokens":64000,"source":"https://aihubmix.com/api/v1/models","supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/mimo-v2-omni":{"mode":"chat","base_model":"mimo-v2-omni","max_input_tokens":256000,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/mimo-v2-pro":{"mode":"chat","base_model":"mimo-v2-pro","max_input_tokens":1000000,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_web_search":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/minimax-m2.7":{"mode":"chat","base_model":"minimax-m2.7","max_input_tokens":204800,"max_output_tokens":204800,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":204800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/minimax-m3":{"mode":"chat","base_model":"minimax-m3","max_input_tokens":1000000,"max_output_tokens":524288,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"aihubmix","max_tokens":524288,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/muse-spark-1.2":{"mode":"chat","base_model":"muse-spark-1.2","max_input_tokens":1048576,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_response_schema":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.5-122b-a10b":{"mode":"chat","base_model":"qwen3.5-122b-a10b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.5-397b-a17b":{"mode":"chat","base_model":"qwen3.5-397b-a17b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.6-27b":{"mode":"chat","base_model":"qwen3.6-27b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.6-35b-a3b":{"mode":"chat","base_model":"qwen3.6-35b-a3b","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.6-max-preview":{"mode":"chat","base_model":"qwen3.6-max","max_input_tokens":262144,"max_output_tokens":65536,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":65536,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.8-2.4t-a95b":{"mode":"chat","base_model":"qwen3.8-2.4t-a95b","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.8-flash":{"mode":"chat","base_model":"qwen3.8-flash","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","max_input_tokens":1000000,"max_output_tokens":131072,"source":"https://aihubmix.com/api/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_web_search":true,"provider":"aihubmix","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"aihubmix/step-3.7-flash":{"mode":"chat","base_model":"step-3.7-flash","max_input_tokens":256000,"source":"https://aihubmix.com/api/v1/models","supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"aihubmix","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/typesafe/jev-1.13":{"mode":"chat","base_model":"jev-1.13","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"supports_vision":true,"supports_decisions":true,"source":"https://openrouter.ai/typesafe/jev-1.13","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":28800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wandb/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1049000,"source":"https://wandb.ai/site/pricing/tokens/","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wandb","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~anthropic/claude-fable-latest":{"mode":"chat","base_model":"claude-fable-latest","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~anthropic/claude-haiku-latest":{"mode":"chat","base_model":"claude-haiku-latest","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~anthropic/claude-opus-latest":{"mode":"chat","base_model":"claude-opus-latest","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~anthropic/claude-sonnet-latest":{"mode":"chat","base_model":"claude-sonnet-latest","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~deepseek/deepseek-flash-latest":{"mode":"chat","base_model":"deepseek-flash-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~deepseek/deepseek-pro-latest":{"mode":"chat","base_model":"deepseek-pro-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":384000,"max_tokens":384000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~deepseek/deepseek-v4-flash-latest":{"mode":"chat","base_model":"deepseek-v4-flash-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~google/gemini-flash-latest":{"mode":"chat","base_model":"gemini-flash-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~google/gemini-pro-latest":{"mode":"chat","base_model":"gemini-pro-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~moonshotai/kimi-latest":{"mode":"chat","base_model":"kimi-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~openai/gpt-astra-latest":{"mode":"chat","base_model":"gpt-astra-latest","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~openai/gpt-luna-latest":{"mode":"chat","base_model":"gpt-luna-latest","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~openai/gpt-mini-latest":{"mode":"chat","base_model":"gpt-mini-latest","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~openai/gpt-sol-latest":{"mode":"chat","base_model":"gpt-sol-latest","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~openai/gpt-terra-latest":{"mode":"chat","base_model":"gpt-terra-latest","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~x-ai/grok-latest":{"mode":"chat","base_model":"grok-latest","provider":"openrouter","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":450000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~z-ai/glm-flash-latest":{"mode":"chat","base_model":"glm-flash-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~z-ai/glm-latest":{"mode":"chat","base_model":"glm-latest","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/aion-labs/aion-2.0":{"mode":"chat","base_model":"aion-2.0","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/aion-labs/aion-3.0":{"mode":"chat","base_model":"aion-3.0","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/aion-labs/aion-3.0-mini":{"mode":"chat","base_model":"aion-3.0-mini","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/amazon/nova-2-lite-v1":{"mode":"chat","base_model":"nova-2-lite-v1","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":65535,"max_tokens":65535,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/amazon/nova-premier-v1":{"mode":"chat","base_model":"nova-premier-v1","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-fable-5:batch":{"mode":"chat","base_model":"claude-fable-5:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-fable-5.1:batch":{"mode":"chat","base_model":"claude-fable-5.1:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-haiku-4.5:batch":{"mode":"chat","base_model":"claude-haiku-4.5:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.1:batch":{"mode":"chat","base_model":"claude-opus-4.1:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.5:batch":{"mode":"chat","base_model":"claude-opus-4.5:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.6:batch":{"mode":"chat","base_model":"claude-opus-4.6:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.7:batch":{"mode":"chat","base_model":"claude-opus-4.7:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-4.8:batch":{"mode":"chat","base_model":"claude-opus-4.8:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-opus-5:batch":{"mode":"chat","base_model":"claude-opus-5:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-sonnet-4.5:batch":{"mode":"chat","base_model":"claude-sonnet-4.5:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-sonnet-4.6:batch":{"mode":"chat","base_model":"claude-sonnet-4.6:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/anthropic/claude-sonnet-5:batch":{"mode":"chat","base_model":"claude-sonnet-5:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/arcee-ai/trinity-large-thinking":{"mode":"chat","base_model":"trinity-large-thinking","max_input_tokens":262144,"max_output_tokens":80000,"max_tokens":80000,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/baidu/ernie-4.5-vl-424b-a47b":{"mode":"chat","base_model":"ernie-4.5-vl-424b-a47b","deprecation_date":"2026-10-08","max_input_tokens":123000,"max_output_tokens":16000,"max_tokens":16000,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true,"is_deprecated":true},"openrouter/bytedance-seed/seed-1.6":{"mode":"chat","base_model":"seed-1.6","deprecation_date":"2026-11-11","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/bytedance-seed/seed-1.6-flash":{"mode":"chat","base_model":"seed-1.6-flash","deprecation_date":"2026-11-11","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/bytedance-seed/seed-2-1-turbo":{"mode":"chat","base_model":"seed-2-1-turbo","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/bytedance-seed/seed-2.0-code":{"mode":"chat","base_model":"seed-2.0-code","deprecation_date":"2026-11-11","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/bytedance-seed/seed-2.0-lite":{"mode":"chat","base_model":"seed-2.0-lite","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/bytedance-seed/seed-2.0-mini":{"mode":"chat","base_model":"seed-2.0-mini","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/cognitivecomputations/dolphin-mistral-24b-venice-edition":{"mode":"chat","base_model":"dolphin-mistral-24b-venice-edition","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/cohere/north-mini-code:free":{"mode":"chat","base_model":"north-mini-code","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/dots-studio/dots-3-note-preview:free":{"mode":"chat","base_model":"dots-studio/dots-3-note","provider":"openrouter","max_input_tokens":512000,"max_output_tokens":460800,"max_tokens":460800,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"deprecation_date":"2026-12-31","source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":460800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-2.5-flash-lite:batch":{"mode":"chat","base_model":"gemini-2.5-flash-lite:batch","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_video_input":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-2.5-flash:batch":{"mode":"chat","base_model":"gemini-2.5-flash:batch","deprecation_date":"2026-10-20","max_input_tokens":1048576,"max_output_tokens":65535,"max_tokens":65535,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_video_input":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/google/gemini-2.5-pro:batch":{"mode":"chat","base_model":"gemini-2.5-pro:batch","deprecation_date":"2026-10-20","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_video_input":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/google/gemini-3-flash-preview:batch":{"mode":"chat","base_model":"gemini-3-flash-preview:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_prompt_caching":false,"supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-lite:batch":{"mode":"chat","base_model":"gemini-3.1-flash-lite:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-pro-preview:batch":{"mode":"chat","base_model":"gemini-3.1-pro-preview:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_prompt_caching":false,"supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.5-flash-lite:batch":{"mode":"chat","base_model":"gemini-3.5-flash-lite:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.5-flash:batch":{"mode":"chat","base_model":"gemini-3.5-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.6-flash:batch":{"mode":"chat","base_model":"gemini-3.6-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.7-flash:batch":{"mode":"chat","base_model":"gemini-3.7-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.8-flash:batch":{"mode":"chat","base_model":"gemini-3.8-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_web_search":true,"supports_video_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/ibm-granite/granite-4.0-h-micro":{"mode":"chat","base_model":"granite-4.0-h-micro","max_input_tokens":131000,"max_output_tokens":117900,"max_tokens":117900,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117900},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/ibm-granite/granite-4.2-8b":{"mode":"chat","base_model":"granite-4.2-8b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inception/mercury-2":{"mode":"chat","base_model":"mercury-2","max_input_tokens":128000,"max_output_tokens":50000,"max_tokens":50000,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":50000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true},"openrouter/inception/mercury-2.5":{"mode":"chat","base_model":"mercury-2.5","max_input_tokens":260000,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true},"openrouter/inclusionai/ling-3.0-flash":{"mode":"chat","base_model":"ling-3.0-flash","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inclusionai/ling-3.0-flash-fin":{"mode":"chat","base_model":"ling-3.0-flash-fin","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inclusionai/ling-3.0-flash-fin:free":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash-fin","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inclusionai/ling-3.0-flash-sante:free":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash-sante","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inclusionai/ling-3.0-flash-vl":{"mode":"chat","base_model":"ling-3.0-flash-vl","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inclusionai/ling-3.0-flash-vl:free":{"mode":"chat","base_model":"inclusionai/ling-3.0-flash-vl","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inference-net/schematron-v2-small":{"mode":"chat","base_model":"schematron-v2-small","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/inference-net/schematron-v2-turbo":{"mode":"chat","base_model":"schematron-v2-turbo","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/kwaipilot/kat-coder-pro-v2.5":{"mode":"chat","base_model":"kat-coder-pro-v2.5","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/liquid/lfm-2.5-2.6b:free":{"mode":"chat","base_model":"liquid/lfm-2.5-2.6b","max_input_tokens":65536,"max_output_tokens":8192,"max_tokens":8192,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meituan/longcat-2.0":{"mode":"chat","base_model":"longcat-2.0","provider":"openrouter","max_input_tokens":1048756,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-glimmer-30b":{"mode":"chat","base_model":"muse-glimmer-30b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-spark-1.1":{"mode":"chat","base_model":"muse-spark-1.1","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-spark-1.2":{"mode":"chat","base_model":"muse-spark-1.2","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-spark-1.2-contributor":{"mode":"chat","base_model":"muse-spark-1.2-contributor","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-spark-1.3":{"mode":"chat","base_model":"muse-spark-1.3","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-spark-1.3-contributor":{"mode":"chat","base_model":"muse-spark-1.3-contributor","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":943718},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/codestral-2508:batch":{"mode":"chat","base_model":"codestral-2508:batch","max_input_tokens":256000,"max_output_tokens":204800,"max_tokens":204800,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/ministral-8b-2512:batch":{"mode":"chat","base_model":"ministral-8b-2512:batch","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-large-2512:batch":{"mode":"chat","base_model":"mistral-large-2512:batch","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-medium-3-5:batch":{"mode":"chat","base_model":"mistral-medium-3-5:batch","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-medium-3.1:batch":{"mode":"chat","base_model":"mistral-medium-3.1:batch","max_input_tokens":131072,"max_output_tokens":104857,"max_tokens":104857,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":104857},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/mistral-small-2603:batch":{"mode":"chat","base_model":"mistral-small-2603:batch","max_input_tokens":262144,"max_output_tokens":209715,"max_tokens":209715,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":209715},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/moonshotai/kimi-k3:batch":{"mode":"chat","base_model":"kimi-k3:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nex-agi/nex-n2.5-mini:free":{"mode":"chat","base_model":"nex-agi/nex-n2.5-mini","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/nex-agi/nex-n2.5-pro:free":{"mode":"chat","base_model":"nex-agi/nex-n2.5-pro","deprecation_date":"2026-09-25","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"is_deprecated":true},"openrouter/nousresearch/hermes-4-405b":{"mode":"chat","base_model":"hermes-4-405b","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":false,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-3.5-turbo:batch":{"mode":"chat","base_model":"gpt-3.5-turbo:batch","max_input_tokens":16385,"max_output_tokens":4096,"max_tokens":4096,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":true,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4-turbo:batch":{"mode":"chat","base_model":"gpt-4-turbo:batch","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":4096,"max_tokens":4096,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4.1-mini:batch":{"mode":"chat","base_model":"gpt-4.1-mini:batch","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4.1-nano:batch":{"mode":"chat","base_model":"gpt-4.1-nano:batch","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4.1:batch":{"mode":"chat","base_model":"gpt-4.1:batch","provider":"openrouter","max_input_tokens":1047576,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4o-mini:batch":{"mode":"chat","base_model":"gpt-4o-mini:batch","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4o:batch":{"mode":"chat","base_model":"gpt-4o:batch","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_reasoning":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-image":{"mode":"chat","base_model":"gpt-5-image","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-image-mini":{"mode":"chat","base_model":"gpt-5-image-mini","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-mini:batch":{"mode":"chat","base_model":"gpt-5-mini:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-nano:batch":{"mode":"chat","base_model":"gpt-5-nano:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5-pro:batch":{"mode":"chat","base_model":"gpt-5-pro:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5:batch":{"mode":"chat","base_model":"gpt-5:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.1:batch":{"mode":"chat","base_model":"gpt-5.1:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.2-pro:batch":{"mode":"chat","base_model":"gpt-5.2-pro:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.2:batch":{"mode":"chat","base_model":"gpt-5.2:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-image-2":{"mode":"chat","base_model":"gpt-5.4-image-2","provider":"openrouter","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-mini:batch":{"mode":"chat","base_model":"gpt-5.4-mini:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-nano:batch":{"mode":"chat","base_model":"gpt-5.4-nano:batch","provider":"openrouter","max_input_tokens":400000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4-pro:batch":{"mode":"chat","base_model":"gpt-5.4-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.4:batch":{"mode":"chat","base_model":"gpt-5.4:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.5-pro:batch":{"mode":"chat","base_model":"gpt-5.5-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_prompt_caching":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.5:batch":{"mode":"chat","base_model":"gpt-5.5:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-luna-pro:batch":{"mode":"chat","base_model":"gpt-5.6-luna-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-luna:batch":{"mode":"chat","base_model":"gpt-5.6-luna:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-sol-pro:batch":{"mode":"chat","base_model":"gpt-5.6-sol-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-sol:batch":{"mode":"chat","base_model":"gpt-5.6-sol:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-terra-pro:batch":{"mode":"chat","base_model":"gpt-5.6-terra-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-5.6-terra:batch":{"mode":"chat","base_model":"gpt-5.6-terra:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-6-astra-pro:batch":{"mode":"chat","base_model":"gpt-6-astra-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-astra:batch":{"mode":"chat","base_model":"gpt-6-astra:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-luna-pro":{"mode":"chat","base_model":"gpt-6-luna-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-sol-pro":{"mode":"chat","base_model":"gpt-6-sol-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/o3-mini:batch":{"mode":"chat","base_model":"o3-mini:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":100000,"max_tokens":100000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_vision":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/o3:batch":{"mode":"chat","base_model":"o3:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":100000,"max_tokens":100000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/o4-mini:batch":{"mode":"chat","base_model":"o4-mini:batch","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":100000,"max_tokens":100000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":100000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/perceptron/perceptron-mk1":{"mode":"chat","base_model":"perceptron-mk1","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":8192,"max_tokens":8192,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/perplexity/sonar-pro-search":{"mode":"chat","base_model":"sonar-pro-search","provider":"openrouter","max_input_tokens":200000,"max_output_tokens":8000,"max_tokens":8000,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-27b:free":{"mode":"chat","base_model":"qwen3.8-27b","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/rekaai/reka-edge":{"mode":"chat","base_model":"reka-edge","provider":"openrouter","max_input_tokens":16384,"max_output_tokens":14745,"max_tokens":14745,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":14745},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/rekaai/reka-flash-3":{"mode":"chat","base_model":"reka-flash-3","provider":"openrouter","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_reasoning":true,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":58982},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/relace/relace-apply-3":{"mode":"chat","base_model":"relace-apply-3","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/relace/relace-search":{"mode":"chat","base_model":"relace-search","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/sakana/fugu-max":{"mode":"chat","base_model":"fugu-max","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/sakana/fugu-ultra":{"mode":"chat","base_model":"fugu-ultra","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/sakana/fugu-ultra-v2":{"mode":"chat","base_model":"fugu-ultra-v2","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/sakana/sakana-namazu":{"mode":"chat","base_model":"sakana-namazu","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/stepfun/step-3.5-flash":{"mode":"chat","base_model":"step-3.5-flash","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/stepfun/step-3.7-flash":{"mode":"chat","base_model":"step-3.7-flash","provider":"openrouter","max_input_tokens":256000,"max_output_tokens":230400,"max_tokens":230400,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":230400},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy-mt2-1.8b":{"mode":"chat","base_model":"hy-mt2-1.8b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy-mt2-30b-a3b":{"mode":"chat","base_model":"hy-mt2-30b-a3b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy-mt2-7b":{"mode":"chat","base_model":"hy-mt2-7b","provider":"openrouter","max_input_tokens":8192,"max_output_tokens":4096,"max_tokens":4096,"supports_response_schema":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy3":{"mode":"chat","base_model":"hy3","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"off_peak_pricing":{"hours_utc":"16:00-00:00","input_cost_per_token":8.25e-8,"output_cost_per_token":3.3e-7,"cache_read_input_token_cost":2.0625e-8},"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy3-preview":{"mode":"chat","base_model":"hy3-preview","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_response_schema":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/tencent/hy4-preview":{"mode":"chat","base_model":"hy4-preview","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/thedrummer/cydonia-24b-v4.1":{"mode":"chat","base_model":"cydonia-24b-v4.1","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_reasoning":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/thinkingmachines/inkling":{"mode":"chat","base_model":"inkling","provider":"openrouter","max_input_tokens":524288,"max_output_tokens":471859,"max_tokens":471859,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_response_schema":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":471859},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/thinkingmachines/inkling-small":{"mode":"chat","base_model":"inkling-small","provider":"openrouter","max_input_tokens":524288,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_response_schema":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/thinkingmachines/inkling-small:free":{"mode":"chat","base_model":"thinkingmachines/inkling-small","max_input_tokens":1048576,"max_output_tokens":262144,"max_tokens":262144,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/thinkingmachines/inkling:free":{"mode":"chat","base_model":"thinkingmachines/inkling","max_input_tokens":1048576,"max_output_tokens":262144,"max_tokens":262144,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/unbiased/pareto":{"mode":"chat","base_model":"pareto","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/upstage/solar-pro-3":{"mode":"chat","base_model":"solar-pro-3","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"supports_web_search":false,"provider":"openrouter","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/upstage/solar-pro4":{"mode":"chat","base_model":"solar-pro4","provider":"openrouter","max_input_tokens":524288,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/writer/palmyra-x5":{"mode":"chat","base_model":"palmyra-x5","provider":"openrouter","max_input_tokens":1040000,"max_output_tokens":8192,"max_tokens":8192,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.3:batch":{"mode":"chat","base_model":"grok-4.3:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":900000,"max_tokens":900000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":900000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.3-flash:batch":{"mode":"chat","base_model":"glm-5.3-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.3:batch":{"mode":"chat","base_model":"glm-5.3:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_vision":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/prism-ml/ternary-bonsai-2-27b":{"mode":"chat","base_model":"ternary-bonsai-2-27b","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":false,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/z-ai/glm-5.3-flashx":{"mode":"chat","base_model":"glm-5.3-flashx","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-4.7":{"mode":"chat","base_model":"grok-4.7","provider":"openrouter","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":450000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/xiaomi/mimo-v2.6-flash":{"mode":"chat","base_model":"mimo-v2.6-flash","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/xiaomi/mimo-v2.6-pro":{"mode":"chat","base_model":"mimo-v2.6-pro","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/xiaomi/mimo-v2.6-pro-ultraspeed":{"mode":"chat","base_model":"mimo-v2.6-pro-ultraspeed","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xiaomi_mimo/mimo-v2.6-pro":{"mode":"chat","base_model":"mimo-v2.6-pro","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.xiaomimimo.com/static/docs/price/pay-as-you-go.md","supported_endpoints":["/v1/chat/completions"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"xiaomi_mimo","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_web_search":true},"xiaomi_mimo/mimo-v2.6-flash":{"mode":"chat","base_model":"mimo-v2.6-flash","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"source":"https://platform.xiaomimimo.com/static/docs/price/pay-as-you-go.md","supported_endpoints":["/v1/chat/completions"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"xiaomi_mimo","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_web_search":true},"xai/grok-4.20-0309":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-0309":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-latest":{"mode":"chat","base_model":"grok-4.20-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-latest-non-reasoning":{"mode":"chat","base_model":"grok-4.20-beta-latest-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-latest-reasoning":{"mode":"chat","base_model":"grok-4.20-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-non-reasoning":{"mode":"chat","base_model":"grok-4.20-beta-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-beta-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-0304":{"mode":"chat","base_model":"grok-4.20-experimental","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-0304-non-reasoning":{"mode":"chat","base_model":"grok-4.20-experimental-beta-0304-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-0304-reasoning":{"mode":"chat","base_model":"grok-4.20-experimental-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-latest":{"mode":"chat","base_model":"grok-4.20-experimental-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-non-reasoning-latest":{"mode":"chat","base_model":"grok-4.20-experimental-beta-non","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-experimental-beta-reasoning-latest":{"mode":"chat","base_model":"grok-4.20-experimental-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-multi-agent-beta-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supported_endpoints":["/v1/responses"],"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-4.20-multi-agent-experimental-beta-0304":{"mode":"responses","base_model":"grok-4.20-multi-agent-experimental","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supported_endpoints":["/v1/responses"],"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-4.20-multi-agent-experimental-beta-latest":{"mode":"responses","base_model":"grok-4.20-multi-agent-experimental-beta","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supported_endpoints":["/v1/responses"],"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/responses"]}]},"xai/grok-4.20-non-reasoning-gv2":{"mode":"chat","base_model":"grok-4.20-non-reasoning-gv2","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"xai/grok-4.20-reasoning-gv2":{"mode":"chat","base_model":"grok-4.20-reasoning-gv2","max_input_tokens":1000000,"max_output_tokens":1000000,"source":"https://api.x.ai/v1/language-models","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"xai","max_tokens":1000000,"supports_audio_input":false,"supports_audio_output":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nex-agi/nex-n2.5-mini":{"mode":"chat","base_model":"nex-n2.5-mini","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_function_calling":false,"supports_pdf_input":false,"supports_tool_choice":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nex-agi/nex-n2.5-pro":{"mode":"chat","base_model":"nex-n2.5-pro","provider":"openrouter","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_pdf_input":false,"supports_web_search":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":235929},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","max_input_tokens":1048576,"max_output_tokens":32768,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"kimi-k2.6","deprecation_date":"2026-10-16","max_input_tokens":262000,"max_output_tokens":262000,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","is_deprecated":true,"max_tokens":262000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"kimi-k2.7-code","deprecation_date":"2026-10-16","max_input_tokens":262000,"max_output_tokens":262000,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","is_deprecated":true,"max_tokens":262000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/moonshotai/Kimi-K3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"mode":"chat","base_model":"nvidia-nemotron-3-ultra-550b-a55b","max_input_tokens":202800,"max_output_tokens":202800,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","max_tokens":202800,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/thinkingmachines/inkling":{"mode":"chat","base_model":"thinkingmachines/inkling","deprecation_date":"2026-10-16","max_input_tokens":1048576,"max_output_tokens":32768,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/thinkingmachines/inkling-small":{"mode":"chat","base_model":"thinkingmachines/inkling-small","deprecation_date":"2026-10-16","max_input_tokens":1048576,"max_output_tokens":32768,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","is_deprecated":true,"max_tokens":32768,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-5.2":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","max_input_tokens":1048576,"max_output_tokens":131072,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1048576,"max_output_tokens":384000,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","max_tokens":384000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":384000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","deprecation_date":"2026-10-16","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","is_deprecated":true,"max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"baseten","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"baseten/zai-org/GLM-5.2-Fast":{"mode":"chat","base_model":"glm-5.2-fast","max_input_tokens":1048576,"max_output_tokens":262144,"source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","max_tokens":262144,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-2.0-flash":{"mode":"chat","base_model":"gemini-2.0-flash","deprecation_date":"2026-06-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/gemini-3.5-flash-lite","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-2.0-flash-lite":{"mode":"chat","base_model":"gemini-2.0-flash-lite","deprecation_date":"2026-06-01","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"vertex_ai/gemini-3.5-flash-lite","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/zai-org/glm-5.2-maas":{"mode":"chat","base_model":"glm-5.2","max_input_tokens":1000000,"max_output_tokens":64000,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_regions":["global"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"vertex_ai","max_tokens":64000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/cohere/command-a-plus":{"mode":"chat","base_model":"command-a-plus","provider":"openrouter","max_input_tokens":192000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-6-luna-pro:batch":{"mode":"chat","base_model":"gpt-6-luna-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-luna:batch":{"mode":"chat","base_model":"gpt-6-luna:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-sol-pro:batch":{"mode":"chat","base_model":"gpt-6-sol-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6-sol:batch":{"mode":"chat","base_model":"gpt-6-sol:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/anthropic/claude-opus-5.5:batch":{"mode":"chat","base_model":"claude-opus-5.5:batch","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/assemblyai/universal-3-5-pro":{"mode":"chat","base_model":"universal-3-5-pro","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3.8-omni-flash":{"mode":"chat","base_model":"qwen3.8-omni-flash","provider":"openrouter","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/~typesafe/jev-latest":{"mode":"chat","base_model":"jev-latest","provider":"openrouter","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"supports_vision":true,"supports_decisions":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":28800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/meta/muse-voice-transcribe-1.0":{"mode":"chat","base_model":"muse-voice-transcribe-1.0","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepseek/deepseek-v4.1-flash:batch":{"mode":"chat","base_model":"deepseek-v4.1-flash:batch","provider":"openrouter","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/microsoft/mai-transcribe-2":{"mode":"chat","base_model":"mai-transcribe-2","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/nemotron-3.5-asr-streaming-multilingual-0.6b":{"mode":"chat","base_model":"nemotron-3.5-asr-streaming-multilingual-0.6b","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/voxtral-small-24b-2507-stt":{"mode":"chat","base_model":"voxtral-small-24b-2507-stt","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/voxtral-mini-3b-2507":{"mode":"chat","base_model":"voxtral-mini-3b-2507","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-asr-1.7b":{"mode":"chat","base_model":"qwen3-asr-1.7b","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-asr-0.6b":{"mode":"chat","base_model":"qwen3-asr-0.6b","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-transcribe":{"mode":"chat","base_model":"gpt-transcribe","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/fish-audio/transcribe-1":{"mode":"chat","base_model":"transcribe-1","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/fish-audio/s1":{"mode":"chat","base_model":"s1","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/fish-audio/s2-pro":{"mode":"chat","base_model":"s2-pro","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/fish-audio/s2.1-pro":{"mode":"chat","base_model":"s2.1-pro","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/microsoft/mai-voice-2-flash":{"mode":"chat","base_model":"mai-voice-2-flash","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen-audio-3.0-tts-flash":{"mode":"chat","base_model":"qwen-audio-3.0-tts-flash","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen-audio-3.0-tts-plus":{"mode":"chat","base_model":"qwen-audio-3.0-tts-plus","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-stt-1.0":{"mode":"chat","base_model":"grok-stt-1.0","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openrouter/auto-beta":{"mode":"chat","base_model":"auto-beta","provider":"openrouter","max_input_tokens":2000000,"max_tokens":2000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepgram/aura-2":{"mode":"chat","base_model":"aura-2","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/speech-2.8-hd":{"mode":"chat","base_model":"speech-2.8-hd","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/minimax/speech-2.8-turbo":{"mode":"chat","base_model":"speech-2.8-turbo","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/deepgram/nova-3":{"mode":"chat","base_model":"nova-3","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openrouter/fusion":{"mode":"chat","base_model":"fusion","provider":"openrouter","max_input_tokens":1000000,"max_tokens":1000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/microsoft/mai-voice-2":{"mode":"chat","base_model":"mai-voice-2","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/microsoft/mai-transcribe-1.5":{"mode":"chat","base_model":"mai-transcribe-1.5","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/nvidia/parakeet-tdt-0.6b-v3":{"mode":"chat","base_model":"parakeet-tdt-0.6b-v3","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/voxtral-mini-transcribe":{"mode":"chat","base_model":"voxtral-mini-transcribe","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/x-ai/grok-voice-tts-1.0":{"mode":"chat","base_model":"grok-voice-tts-1.0","provider":"openrouter","max_input_tokens":15000,"max_output_tokens":13500,"max_tokens":13500,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":13500},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/qwen/qwen3-asr-flash-2026-02-10":{"mode":"chat","base_model":"qwen3-asr-flash-2026-02-10","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/chirp-3":{"mode":"chat","base_model":"chirp-3","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4o-mini-transcribe":{"mode":"chat","base_model":"gpt-4o-mini-transcribe","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":115200,"max_tokens":115200,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":115200},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/whisper-large-v3":{"mode":"chat","base_model":"whisper-large-v3","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/whisper-large-v3-turbo":{"mode":"chat","base_model":"whisper-large-v3-turbo","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/whisper-1":{"mode":"chat","base_model":"whisper-1","provider":"openrouter","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-4o-transcribe":{"mode":"chat","base_model":"gpt-4o-transcribe","provider":"openrouter","max_input_tokens":128000,"max_output_tokens":115200,"max_tokens":115200,"supports_audio_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":115200},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/google/gemini-3.1-flash-tts-preview":{"mode":"chat","base_model":"gemini-3.1-flash-tts-preview","provider":"openrouter","max_input_tokens":32768,"max_output_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/canopylabs/orpheus-3b-0.1-ft":{"mode":"chat","base_model":"orpheus-3b-0.1-ft","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3686,"range":{"min":1,"max":3686},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/sesame/csm-1b":{"mode":"chat","base_model":"csm-1b","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3686,"range":{"min":1,"max":3686},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/hexgrad/kokoro-82m":{"mode":"chat","base_model":"kokoro-82m","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3686,"range":{"min":1,"max":3686},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openrouter/pareto-code":{"mode":"chat","base_model":"pareto-code","provider":"openrouter","max_input_tokens":2000000,"max_tokens":2000000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/mistralai/voxtral-mini-tts-2603":{"mode":"chat","base_model":"voxtral-mini-tts-2603","provider":"openrouter","max_input_tokens":4096,"max_output_tokens":3276,"max_tokens":3276,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":3276,"range":{"min":1,"max":3276},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"openrouter/openai/gpt-oss-20b:batch":{"mode":"chat","base_model":"gpt-oss-20b:batch","provider":"openrouter","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":117964},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/xai/grok-4.20-beta-0309-non-reasoning":{"mode":"chat","base_model":"grok-4.20","max_input_tokens":2000000,"max_output_tokens":2000000,"max_tokens":2000000,"source":"https://docs.x.ai/docs/models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":2000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/moonshotai.kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"source":"https://platform.kimi.ai/docs/guide/kimi-k2-6-quickstart","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"bedrock","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"moonshotai.kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"moonshot","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemma-4-31b":{"mode":"chat","base_model":"gemma-4","max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":250000,"rpm":10,"provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemma-4-26b-a4b-it-maas":{"mode":"chat","base_model":"gemma-4","max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"tpm":250000,"rpm":10,"provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/DeepSeek-V4-Flash":{"mode":"chat","base_model":"deepseek-v4-flash","provider":"azure","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"azure/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","provider":"azure","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions"]}]},"moonshot/kimi-k2.7-code-highspeed":{"mode":"chat","base_model":"kimi-k2.7-code-highspeed","max_input_tokens":262144,"max_tokens":262144,"supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_vision":true,"provider":"moonshot","model":"kimi-k2.7-code-highspeed","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepseek-v4-pro-0813":{"mode":"chat","base_model":"deepseek-v4-pro","max_input_tokens":1000000,"max_output_tokens":393216,"max_tokens":393216,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"deepseek","peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"supports_native_effort":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"deepseek/deepSeek-v4-flash-0731":{"mode":"chat","base_model":"deepseek-v4-flash","max_input_tokens":1000000,"max_output_tokens":393216,"max_tokens":393216,"source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"deepseek","peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":393216},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_native_effort":true},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-safeguard-120b":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-safeguard-20b":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-5.6-terra":{"mode":"responses","base_model":"gpt-5.6-terra","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-5.6-luna":{"mode":"responses","base_model":"gpt-5.6-luna","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-west-1/anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","provider":"bedrock_mantle","supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":true,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","provider":"bedrock_mantle","supports_prompt_caching":true,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","provider":"bedrock_mantle","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_input_tokens":200000,"max_output_tokens":4000,"max_tokens":4000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-east-1/anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","provider":"bedrock_mantle","supports_safeguards":true,"supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-sonnet-4-5-20250929-v1:0":{"mode":"chat","base_model":"claude-sonnet-4-5","provider":"bedrock_mantle","supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_reasoning":true,"supports_response_schema":true,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-3-7-sonnet-20250219-v1:0":{"mode":"chat","base_model":"claude-3-7-sonnet","provider":"bedrock_mantle","supports_prompt_caching":true,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-3-5-sonnet-20240620-v1:0":{"mode":"chat","base_model":"claude-3-5-sonnet","provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","provider":"bedrock_mantle","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","ca-central-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-south-1","ap-southeast-1","ap-southeast-2","sa-east-1"],"max_input_tokens":200000,"max_output_tokens":4000,"max_tokens":4000,"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-oss-20b-1:0":{"mode":"chat","base_model":"gpt-oss-20b","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/openai.gpt-oss-120b-1:0":{"mode":"chat","base_model":"gpt-oss-120b","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nemotron-nano-9b-v2","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"deprecation_date":"2027-03-30","is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nemotron-nano-12b-v2","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"deprecation_date":"2027-03-30","is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nemotron-nano-3-30b","max_input_tokens":256000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supports_reasoning":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-east-1/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-southeast-2","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nemotron-nano-9b-v2","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"deprecation_date":"2027-03-30","is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nemotron-nano-12b-v2","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"deprecation_date":"2027-03-30","is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nemotron-nano-3-30b","max_input_tokens":256000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supports_reasoning":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/batch"],"supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-southeast-2","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/nvidia.nemotron-nano-9b-v2":{"mode":"chat","base_model":"nemotron-nano-9b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/nvidia.nemotron-nano-12b-v2":{"mode":"chat","base_model":"nemotron-nano-12b-v2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/nvidia.nemotron-nano-3-30b":{"mode":"chat","base_model":"nemotron-nano-3-30b","max_input_tokens":256000,"max_output_tokens":8000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_reasoning":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","us-gov-west-1","eu-central-1","eu-north-1","us-gov-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/nvidia.nemotron-super-3-120b":{"mode":"chat","base_model":"nemotron-super-3-120b","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions","/v1/batch"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-southeast-2","sa-east-1","ap-southeast-4","us-gov-west-1","us-gov-east-1"],"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/anthropic.claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"provider":"bedrock_mantle","supports_forced_tool_choice":false,"supports_web_search":true,"tool_use_system_prompt_tokens":346,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_safeguards":true,"supports_mid_conversation_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"supports_web_search":true,"provider":"bedrock","model":"kimi-k3","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"min_output_tokens":16,"source":"https://aws.amazon.com/bedrock/pricing/","max_output_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/global.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"provider":"bedrock","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_video_input":false,"supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_native_streaming":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3.html","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us.moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","max_input_tokens":1000000,"max_output_tokens":1048576,"max_tokens":1000000,"provider":"bedrock","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_video_input":false,"supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_native_streaming":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"source":"https://aws.amazon.com/bedrock/pricing/","min_output_tokens":16,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"us-gov.anthropic.claude-3-haiku-20240307-v1:0":{"mode":"chat","base_model":"claude-3-haiku","deprecation_date":"2026-09-10","max_input_tokens":200000,"max_output_tokens":4096,"max_tokens":4096,"supports_cache_point":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","source":"https://aws.amazon.com/bedrock/pricing/","is_deprecated":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":4096},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/anthropic.claude-opus-5-5":{"mode":"chat","base_model":"anthropic.claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","thinking_always_on":true,"supports_forced_tool_use":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"default_reasoning_effort":"medium","supports_image_input":true,"supports_parallel_function_calling":true,"supports_forced_tool_choice":false,"supports_system_messages":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_safeguards":true,"supports_mid_conversation_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/global.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","provider":"bedrock","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"regional_endpoint_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_image_input":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_none_reasoning_effort":true,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"source":"https://developers.openai.com/api/docs/models/gpt-6-sol","metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/global.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://developers.openai.com/api/docs/models/gpt-6-sol","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"provider":"bedrock_mantle","regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_parallel_function_calling":true,"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/global.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://developers.openai.com/api/docs/models/gpt-6-luna","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"provider":"bedrock","regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_parallel_function_calling":true,"supports_image_input":true,"supports_reasoning_with_tool_calls":true,"supports_low_reasoning_effort":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/global.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","provider":"bedrock_mantle","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"regional_endpoint_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_image_input":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":false,"supports_none_reasoning_effort":true,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"source":"https://developers.openai.com/api/docs/models/gpt-6-luna","metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","thinking_always_on":true,"supports_forced_tool_use":false,"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supports_web_search":true,"provider":"bedrock_mantle","regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_parallel_function_calling":true,"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","provider":"bedrock","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"regional_endpoint_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_image_input":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_none_reasoning_effort":true,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gemini/gemini-robotics-er-1.6-preview":{"mode":"chat","base_model":"gemini-robotics-er-1.6-preview","provider":"gemini","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":1064000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/gemini-3.1-flash-lite-preview":{"mode":"chat","base_model":"gemini-3.1-flash-lite-preview","deprecation_date":"2026-05-25","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.1-flash-lite","max_input_tokens":1000000,"max_output_tokens":64000,"rpm":15,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"tpm":250000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"gemini","is_deprecated":true,"max_tokens":1064000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/deep-research-preview-04-2026":{"mode":"chat","base_model":"deep-research-preview-04-2026","max_input_tokens":1000000,"max_output_tokens":64000,"rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"provider":"gemini","model":"deep-research-preview-04-2026","max_tokens":1064000,"web_search_billing_unit":"per_query","supports_url_context":true,"supports_code_execution":true,"supports_file_search":true,"supports_native_structured_output":false,"supports_native_streaming":true,"supports_reasoning":true,"supports_image_input":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"supports_multimodal":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/deep-research-max-preview-04-2026":{"mode":"chat","base_model":"deep-research-max-preview-04-2026","max_input_tokens":1000000,"max_output_tokens":64000,"rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"provider":"gemini","model":"deep-research-max-preview-04-2026","max_tokens":1064000,"web_search_billing_unit":"per_query","supports_url_context":true,"supports_code_execution":true,"supports_file_search":true,"supports_native_structured_output":false,"supports_native_streaming":true,"supports_reasoning":true,"supports_image_input":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"supports_multimodal":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini/antigravity-preview-05-2026":{"mode":"chat","base_model":"antigravity-preview-05-2026","provider":"gemini","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":1064000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/grok-4.6":{"mode":"chat","base_model":"grok-4.6","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/glm-5.3-flash":{"mode":"chat","base_model":"glm-5.3-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/glm-5.3":{"mode":"chat","base_model":"glm-5.3","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/glm-5.2":{"mode":"chat","base_model":"glm-5.2","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/glm-5.1":{"mode":"chat","base_model":"glm-5.1","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/kimi-k3":{"mode":"chat","base_model":"kimi-k3","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/kimi-k2.7-code":{"mode":"chat","base_model":"kimi-k2.7-code","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":262144,"max_input_tokens":262144,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/kimi-k2.6":{"mode":"chat","base_model":"kimi-k2.6","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/longcat-2.0":{"mode":"chat","base_model":"longcat-2.0","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/deepseek-v4.1-flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/deepseek-v4-pro":{"mode":"chat","base_model":"deepseek-v4-pro","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":384000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/deepseek-v4-flash":{"mode":"chat","base_model":"deepseek-v4-flash","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_native_effort":true},"opencode-zen/deepseek-v4-flash-vision-exp":{"mode":"chat","base_model":"deepseek-v4-flash-vision-exp","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_effort":true},"opencode-go/mimo-v2.5":{"mode":"chat","base_model":"mimo-v2.5","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":128000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/mimo-v2.5-pro":{"mode":"chat","base_model":"mimo-v2.5-pro","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":128000,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/minimax-m3":{"mode":"chat","base_model":"minimax-m3","supported_endpoints":["/v1/messages"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/minimax-m2.7":{"mode":"chat","base_model":"minimax-m2.7","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/muse-spark-1.3-contributor":{"mode":"chat","base_model":"muse-spark-1.3-contributor","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/muse-spark-1.2-contributor":{"mode":"chat","base_model":"muse-spark-1.2-contributor","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/qwen3.8-max":{"mode":"chat","base_model":"qwen3.8-max","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/qwen3.8-flash":{"mode":"chat","base_model":"qwen3.8-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/qwen3.7-max":{"mode":"chat","base_model":"qwen3.7-max","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true},"opencode-go/qwen3.7-plus":{"mode":"chat","base_model":"qwen3.7-plus","supported_endpoints":["/v1/messages"],"max_output_tokens":65536,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"tiered_pricing":[{"range":[0,256000],"input_cost_per_token":4.0000000000000003e-7,"output_cost_per_token":0.0000016000000000000001,"cache_read_input_token_cost":4e-8,"cache_creation_input_token_cost":5e-7},{"range":[256000,1000000],"input_cost_per_token":0.0000012,"output_cost_per_token":0.0000048,"cache_read_input_token_cost":1.2e-7,"cache_creation_input_token_cost":0.0000015}],"source":"https://opencode.ai/docs/go/","provider":"opencode-go","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-go/qwen3.6-plus":{"mode":"chat","base_model":"qwen3.6-plus","provider":"opencode-go","source":"https://opencode.ai/docs/go/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/hy4-preview":{"mode":"chat","base_model":"hy4-preview","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/hy3":{"mode":"chat","base_model":"hy3","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-6-astra":{"mode":"chat","base_model":"gpt-6-astra","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","max_input_tokens":1050000,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_prompt_caching":true,"supports_native_streaming":true,"supports_system_messages":true,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"opencode-zen/gpt-5.6-sol":{"mode":"chat","base_model":"gpt-5.6-sol","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.5":{"mode":"chat","base_model":"gpt-5.5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.5-pro":{"mode":"chat","base_model":"gpt-5.5-pro","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.4":{"mode":"chat","base_model":"gpt-5.4","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.4-pro":{"mode":"chat","base_model":"gpt-5.4-pro","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.4-mini":{"mode":"chat","base_model":"gpt-5.4-mini","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.4-nano":{"mode":"chat","base_model":"gpt-5.4-nano","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.3-codex":{"mode":"chat","base_model":"gpt-5.3-codex","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.3-codex-spark":{"mode":"chat","base_model":"gpt-5.3-codex-spark","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.2":{"mode":"chat","base_model":"gpt-5.2","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.2-codex":{"mode":"chat","base_model":"gpt-5.2-codex","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.1":{"mode":"chat","base_model":"gpt-5.1","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.1-codex":{"mode":"chat","base_model":"gpt-5.1-codex","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.1-codex-max":{"mode":"chat","base_model":"gpt-5.1-codex-max","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5.1-codex-mini":{"mode":"chat","base_model":"gpt-5.1-codex-mini","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5":{"mode":"chat","base_model":"gpt-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5-codex":{"mode":"chat","base_model":"gpt-5-codex","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gpt-5-nano":{"mode":"chat","base_model":"gpt-5-nano","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-fable-5-1":{"mode":"chat","base_model":"claude-fable-5-1","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-opus-4-5":{"mode":"chat","base_model":"claude-opus-4-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.8-flash":{"mode":"chat","base_model":"gemini-3.8-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.6-flash":{"mode":"chat","base_model":"gemini-3.6-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.5-flash":{"mode":"chat","base_model":"gemini-3.5-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.5-flash-lite":{"mode":"chat","base_model":"gemini-3.5-flash-lite","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3.1-pro":{"mode":"chat","base_model":"gemini-3.1-pro","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/gemini-3-flash":{"mode":"chat","base_model":"gemini-3-flash","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/grok-4.5":{"mode":"chat","base_model":"grok-4.5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/grok-build-0.1":{"mode":"chat","base_model":"grok-build-0.1","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/muse-spark-1.3":{"mode":"chat","base_model":"muse-spark-1.3","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/muse-spark-1.2":{"mode":"chat","base_model":"muse-spark-1.2","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/qwen3.5-plus":{"mode":"chat","base_model":"qwen3.5-plus","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/glm-5":{"mode":"chat","base_model":"glm-5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/kimi-k2.5":{"mode":"chat","base_model":"kimi-k2.5","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/big-pickle":{"mode":"chat","base_model":"big-pickle","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/mimo-v2.5-free":{"mode":"chat","base_model":"mimo-v2.5-free","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/ling-3.0-flash-fin-free":{"mode":"chat","base_model":"ling-3.0-flash-fin-free","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/nemotron-3-ultra-free":{"mode":"chat","base_model":"nemotron-3-ultra-free","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/nemotron-3.5-lightning-free":{"mode":"chat","base_model":"nemotron-3.5-lightning-free","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"opencode-zen/muse-spark-1.3-contributor-free":{"mode":"chat","base_model":"muse-spark-1.3-contributor-free","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"DeepSeek-V4.1-Flash","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/deepseek-ai/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"DeepSeek-V4.1-Flash","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"DeepSeek-V4.1-Flash","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.8-27B":{"mode":"chat","base_model":"Qwen3.8-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/cerebras/Qwen/Qwen3.8-27B":{"mode":"chat","base_model":"Qwen3.8-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.8-27B":{"mode":"chat","base_model":"Qwen3.8-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"GLM-5.3-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"GLM-5.3-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048575,"max_tokens":1048575,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048575},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"GLM-5.3-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-5.3-Flash":{"mode":"chat","base_model":"GLM-5.3-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/meta-llama/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"Llama-3.1-8B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":16384,"max_tokens":16384,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/meta-llama/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"Llama-3.1-8B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/meta-llama/Llama-3.1-8B-Instruct":{"mode":"chat","base_model":"Llama-3.1-8B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/moonshotai/Kimi-K3":{"mode":"chat","base_model":"Kimi-K3","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/moonshotai/Kimi-K3":{"mode":"chat","base_model":"Kimi-K3","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/moonshotai/Kimi-K3":{"mode":"chat","base_model":"Kimi-K3","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/moonshotai/Kimi-K3":{"mode":"chat","base_model":"Kimi-K3","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-5.3":{"mode":"chat","base_model":"GLM-5.3","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/zai-org/GLM-5.3":{"mode":"chat","base_model":"GLM-5.3","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048575,"max_tokens":1048575,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048575},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/zai-org/GLM-5.3":{"mode":"chat","base_model":"GLM-5.3","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-5.3":{"mode":"chat","base_model":"GLM-5.3","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31B-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/google/gemma-4-31B-it":{"mode":"chat","base_model":"gemma-4-31B-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/Qwen/Qwen3.5-9B":{"mode":"chat","base_model":"Qwen3.5-9B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-9B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/ovhcloud/Qwen/Qwen3.5-9B":{"mode":"chat","base_model":"Qwen3.5-9B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-9B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.5-9B":{"mode":"chat","base_model":"Qwen3.5-9B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-9B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/scaleway/Qwen/Qwen3.6-35B-A3B":{"mode":"chat","base_model":"Qwen3.6-35B-A3B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/Qwen/Qwen3.6-35B-A3B":{"mode":"chat","base_model":"Qwen3.6-35B-A3B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/groq/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/cerebras/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_web_search":true},"huggingface/together/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_web_search":true},"huggingface/scaleway/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":128072,"max_tokens":128072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/ovhcloud/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/openai/gpt-oss-120b":{"mode":"chat","base_model":"gpt-oss-120b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen3-8B":{"mode":"chat","base_model":"Qwen3-8B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-8B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/meta-models/Muse-Glimmer-30B":{"mode":"chat","base_model":"Muse-Glimmer-30B","provider":"huggingface","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/meta-models/Muse-Glimmer-30B":{"mode":"chat","base_model":"Muse-Glimmer-30B","provider":"huggingface","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-R1":{"mode":"chat","base_model":"DeepSeek-R1","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1","max_input_tokens":64000,"max_tokens":64000,"supports_response_schema":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.8-2.4T-A95B":{"mode":"chat","base_model":"Qwen3.8-2.4T-A95B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/Qwen/Qwen3.8-2.4T-A95B":{"mode":"chat","base_model":"Qwen3.8-2.4T-A95B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","max_input_tokens":1010000,"max_tokens":1010000,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1010000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false},"huggingface/deepinfra/Qwen/Qwen3.8-2.4T-A95B":{"mode":"chat","base_model":"Qwen3.8-2.4T-A95B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/groq/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_tool_choice":true,"supports_vision":false,"supports_system_messages":true,"supports_reasoning":true},"huggingface/nscale/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_web_search":true},"huggingface/ovhcloud/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/openai/gpt-oss-20b":{"mode":"chat","base_model":"gpt-oss-20b","provider":"huggingface","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true},"huggingface/baseten/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-Coder-Next":{"mode":"chat","base_model":"Qwen3-Coder-Next","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-Coder-Next","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/novita/google/gemma-4-26B-A4B-it":{"mode":"chat","base_model":"gemma-4-26B-A4B-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/scaleway/google/gemma-4-26B-A4B-it":{"mode":"chat","base_model":"gemma-4-26B-A4B-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/google/gemma-4-26B-A4B-it":{"mode":"chat","base_model":"gemma-4-26B-A4B-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"DeepSeek-V4-Pro","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"DeepSeek-V4-Pro","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4-Pro":{"mode":"chat","base_model":"DeepSeek-V4-Pro","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"DeepSeek-V4-Flash","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":false},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4-Flash":{"mode":"chat","base_model":"DeepSeek-V4-Flash","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/tencent/Hy4-preview":{"mode":"chat","base_model":"Hy4-preview","provider":"huggingface","source":"https://huggingface.co/tencent/Hy4-preview","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.6-27B":{"mode":"chat","base_model":"Qwen3.6-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.6-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"mode":"chat","base_model":"DeepSeek-V4-Flash-Vision-Exp","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"mode":"chat","base_model":"DeepSeek-V4-Flash-Vision-Exp","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/inclusionAI/Ling-3.0-flash-Fin":{"mode":"chat","base_model":"Ling-3.0-flash-Fin","provider":"huggingface","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"MiniMax-M3","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"MiniMax-M3","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true},"huggingface/fireworks-ai/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"MiniMax-M3","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":512000,"max_tokens":512000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":512000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/MiniMaxAI/MiniMax-M3":{"mode":"chat","base_model":"MiniMax-M3","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen3-4B-Instruct-2507":{"mode":"chat","base_model":"Qwen3-4B-Instruct-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"Llama-3.3-70B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":12288,"max_tokens":12288,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":12288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"Llama-3.3-70B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_native_streaming":true,"supports_reasoning":false,"supports_vision":false},"huggingface/scaleway/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"Llama-3.3-70B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_native_streaming":true,"supports_reasoning":false,"supports_vision":false},"huggingface/ovhcloud/meta-llama/Llama-3.3-70B-Instruct":{"mode":"chat","base_model":"Llama-3.3-70B-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_native_streaming":true,"supports_reasoning":false,"supports_vision":false},"huggingface/deepinfra/google/gemma-3-4b-it":{"mode":"chat","base_model":"gemma-3-4b-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-3-4b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/inclusionAI/Ling-3.0-flash-VL":{"mode":"chat","base_model":"Ling-3.0-flash-VL","provider":"huggingface","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-VL","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/scaleway/Qwen/Qwen3-Coder-30B-A3B-Instruct":{"mode":"chat","base_model":"Qwen3-Coder-30B-A3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/thinkingmachines/Inkling-Small":{"mode":"chat","base_model":"Inkling-Small","provider":"huggingface","source":"https://huggingface.co/thinkingmachines/Inkling-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/thinkingmachines/Inkling-Small":{"mode":"chat","base_model":"Inkling-Small","provider":"huggingface","source":"https://huggingface.co/thinkingmachines/Inkling-Small","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048575,"max_tokens":1048575,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048575},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true,"supports_parallel_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false},"huggingface/scaleway/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true,"supports_parallel_function_calling":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false},"huggingface/baseten/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/thinkingmachines/Inkling":{"mode":"chat","base_model":"Inkling","provider":"huggingface","source":"https://huggingface.co/thinkingmachines/Inkling","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/thinkingmachines/Inkling":{"mode":"chat","base_model":"Inkling","provider":"huggingface","source":"https://huggingface.co/thinkingmachines/Inkling","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_prompt_caching":true},"huggingface/deepinfra/thinkingmachines/Inkling":{"mode":"chat","base_model":"Inkling","provider":"huggingface","source":"https://huggingface.co/thinkingmachines/Inkling","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":524288},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/swiss-ai/Apertus-v1.5-70B":{"mode":"chat","base_model":"Apertus-v1.5-70B","provider":"huggingface","source":"https://huggingface.co/swiss-ai/Apertus-v1.5-70B","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"DeepSeek-V4-Pro-0813","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/together/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"DeepSeek-V4-Pro-0813","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"DeepSeek-V4-Pro-0813","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813":{"mode":"chat","base_model":"DeepSeek-V4-Pro-0813","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/XiaomiMiMo/MiMo-V2.5-Pro":{"mode":"chat","base_model":"MiMo-V2.5-Pro","provider":"huggingface","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false},"huggingface/deepinfra/XiaomiMiMo/MiMo-V2.5-Pro":{"mode":"chat","base_model":"MiMo-V2.5-Pro","provider":"huggingface","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.7-Flash":{"mode":"chat","base_model":"GLM-4.7-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.7-Flash","max_input_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-4.7-Flash":{"mode":"chat","base_model":"GLM-4.7-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.7-Flash","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/inclusionAI/Ling-3.0-flash":{"mode":"chat","base_model":"Ling-3.0-flash","provider":"huggingface","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/inclusionAI/Ling-3.0-flash":{"mode":"chat","base_model":"Ling-3.0-flash","provider":"huggingface","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"Kimi-K2.7-Code","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"Kimi-K2.7-Code","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/baseten/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"Kimi-K2.7-Code","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262000,"max_tokens":262000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/moonshotai/Kimi-K2.7-Code":{"mode":"chat","base_model":"Kimi-K2.7-Code","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/microsoft/phi-4":{"mode":"chat","base_model":"phi-4","provider":"huggingface","source":"https://huggingface.co/microsoft/phi-4","max_input_tokens":16384,"max_tokens":16384,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"Llama-4-Scout-17B-16E-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"Llama-4-Scout-17B-16E-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct","max_input_tokens":890000,"max_tokens":890000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":890000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"mode":"chat","base_model":"Llama-4-Scout-17B-16E-Instruct","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct","max_input_tokens":327680,"max_tokens":327680,"supports_response_schema":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":327680},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen2.5-Coder-32B-Instruct":{"mode":"chat","base_model":"Qwen2.5-Coder-32B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.5-27B":{"mode":"chat","base_model":"Qwen3.5-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.5-27B":{"mode":"chat","base_model":"Qwen3.5-27B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/XiaomiMiMo/MiMo-V2.5":{"mode":"chat","base_model":"MiMo-V2.5","provider":"huggingface","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true},"huggingface/deepinfra/XiaomiMiMo/MiMo-V2.5":{"mode":"chat","base_model":"MiMo-V2.5","provider":"huggingface","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3-30B-A3B":{"mode":"chat","base_model":"Qwen3-30B-A3B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-30B-A3B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"Kimi-K2.5","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.5","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/moonshotai/Kimi-K2.5":{"mode":"chat","base_model":"Kimi-K2.5","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.5","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4":{"mode":"chat","base_model":"NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4","provider":"huggingface","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.5-35B-A3B":{"mode":"chat","base_model":"Qwen3.5-35B-A3B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.5-35B-A3B":{"mode":"chat","base_model":"Qwen3.5-35B-A3B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Sao10K/L3-8B-Stheno-v3.2":{"mode":"chat","base_model":"L3-8B-Stheno-v3.2","provider":"huggingface","source":"https://huggingface.co/Sao10K/L3-8B-Stheno-v3.2","max_input_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/google/gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-3-12b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/speakleash/Bielik-11B-v3.0-Instruct":{"mode":"chat","base_model":"Bielik-11B-v3.0-Instruct","provider":"huggingface","source":"https://huggingface.co/speakleash/Bielik-11B-v3.0-Instruct","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen3-14B":{"mode":"chat","base_model":"Qwen3-14B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-14B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/Qwen/Qwen3-14B":{"mode":"chat","base_model":"Qwen3-14B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-14B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen2.5-Coder-7B-Instruct":{"mode":"chat","base_model":"Qwen2.5-Coder-7B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/moonshotai/Kimi-K2-Instruct":{"mode":"chat","base_model":"Kimi-K2-Instruct","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/google/gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","provider":"huggingface","source":"https://huggingface.co/google/gemma-3-27b-it","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V3.2":{"mode":"chat","base_model":"DeepSeek-V3.2","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_vision":false},"huggingface/deepinfra/deepseek-ai/DeepSeek-V3.2":{"mode":"chat","base_model":"DeepSeek-V3.2","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B":{"mode":"chat","base_model":"DeepSeek-R1-Distill-Qwen-7B","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B","max_input_tokens":131072,"max_tokens":131072,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-VL-235B-A22B-Instruct":{"mode":"chat","base_model":"Qwen3-VL-235B-A22B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true},"huggingface/deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct":{"mode":"chat","base_model":"Qwen3-VL-235B-A22B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16":{"mode":"chat","base_model":"NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","provider":"huggingface","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"Kimi-K2.6","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/fireworks-ai/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"Kimi-K2.6","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/baseten/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"Kimi-K2.6","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262000,"max_tokens":262000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/deepinfra/moonshotai/Kimi-K2.6":{"mode":"chat","base_model":"Kimi-K2.6","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-VL-30B-A3B-Instruct":{"mode":"chat","base_model":"Qwen3-VL-30B-A3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-VL-30B-A3B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct":{"mode":"chat","base_model":"Qwen3-VL-30B-A3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-VL-30B-A3B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/tencent/Hy3":{"mode":"chat","base_model":"Hy3","provider":"huggingface","source":"https://huggingface.co/tencent/Hy3","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/stepfun-ai/Step-3.7-Flash":{"mode":"chat","base_model":"Step-3.7-Flash","provider":"huggingface","source":"https://huggingface.co/stepfun-ai/Step-3.7-Flash","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/ibm-granite/granite-4.2-3b":{"mode":"chat","base_model":"granite-4.2-3b","provider":"huggingface","source":"https://huggingface.co/ibm-granite/granite-4.2-3b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"mode":"chat","base_model":"Llama-4-Maverick-17B-128E-Instruct-FP8","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","max_input_tokens":1048576,"max_tokens":1048576,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-R1-Distill-Llama-70B":{"mode":"chat","base_model":"DeepSeek-R1-Distill-Llama-70B","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B","max_input_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true},"huggingface/deepinfra/meta-llama/Llama-Guard-4-12B":{"mode":"chat","base_model":"Llama-Guard-4-12B","provider":"huggingface","source":"https://huggingface.co/meta-llama/Llama-Guard-4-12B","max_input_tokens":163840,"max_tokens":163840,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B":{"mode":"chat","base_model":"DeepSeek-R1-Distill-Qwen-14B","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.5-122B-A10B":{"mode":"chat","base_model":"Qwen3.5-122B-A10B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-122B-A10B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.5-122B-A10B":{"mode":"chat","base_model":"Qwen3.5-122B-A10B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-122B-A10B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"mode":"chat","base_model":"Qwen3-Coder-480B-A35B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/novita/zai-org/GLM-5":{"mode":"chat","base_model":"GLM-5","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":202800,"max_tokens":202800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-5":{"mode":"chat","base_model":"GLM-5","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M2.5":{"mode":"chat","base_model":"MiniMax-M2.5","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.5","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V3":{"mode":"chat","base_model":"DeepSeek-V3","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V3-0324":{"mode":"chat","base_model":"DeepSeek-V3-0324","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3-0324","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/ibm-granite/granite-4.2-8b":{"mode":"chat","base_model":"granite-4.2-8b","provider":"huggingface","source":"https://huggingface.co/ibm-granite/granite-4.2-8b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/ibm-granite/granite-4.2-30b":{"mode":"chat","base_model":"granite-4.2-30b","provider":"huggingface","source":"https://huggingface.co/ibm-granite/granite-4.2-30b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen2.5-Coder-3B-Instruct":{"mode":"chat","base_model":"Qwen2.5-Coder-3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-Coder-3B-Instruct","max_input_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"Qwen3.5-397B-A17B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/scaleway/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"Qwen3.5-397B-A17B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":true},"huggingface/ovhcloud/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"Qwen3.5-397B-A17B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/Qwen/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"Qwen3.5-397B-A17B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Sao10K/L3-8B-Lunaris-v1":{"mode":"chat","base_model":"L3-8B-Lunaris-v1","provider":"huggingface","source":"https://huggingface.co/Sao10K/L3-8B-Lunaris-v1","max_input_tokens":8192,"max_tokens":8192,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/novita/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"Qwen2.5-72B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-72B-Instruct","max_input_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/Qwen/Qwen2.5-72B-Instruct":{"mode":"chat","base_model":"Qwen2.5-72B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-72B-Instruct","max_input_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen3-32B":{"mode":"chat","base_model":"Qwen3-32B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-32B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/Qwen/Qwen3-32B":{"mode":"chat","base_model":"Qwen3-32B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-32B","max_input_tokens":40960,"max_tokens":40960,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/moonshotai/Kimi-K2-Instruct-0905":{"mode":"chat","base_model":"Kimi-K2-Instruct-0905","provider":"huggingface","source":"https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-Next-80B-A3B-Instruct":{"mode":"chat","base_model":"Qwen3-Next-80B-A3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct":{"mode":"chat","base_model":"Qwen3-Next-80B-A3B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/Qwen/Qwen3-4B-Thinking-2507":{"mode":"chat","base_model":"Qwen3-4B-Thinking-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M2.7":{"mode":"chat","base_model":"MiniMax-M2.7","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.7","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/MiniMaxAI/MiniMax-M2.7":{"mode":"chat","base_model":"MiniMax-M2.7","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.7","max_input_tokens":196608,"max_tokens":196608,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":196608},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/stepfun-ai/Step-3.5-Flash":{"mode":"chat","base_model":"Step-3.5-Flash","provider":"huggingface","source":"https://huggingface.co/stepfun-ai/Step-3.5-Flash","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Instruct-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"huggingface/nscale/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Instruct-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","max_input_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"huggingface/scaleway/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Instruct-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"huggingface/deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Instruct-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-VL-235B-A22B-Thinking":{"mode":"chat","base_model":"Qwen3-VL-235B-A22B-Thinking","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Thinking","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.5-Air":{"mode":"chat","base_model":"GLM-4.5-Air","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.5-Air","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/AutoGLM-Phone-9B-Multilingual":{"mode":"chat","base_model":"AutoGLM-Phone-9B-Multilingual","provider":"huggingface","source":"https://huggingface.co/zai-org/AutoGLM-Phone-9B-Multilingual","max_input_tokens":65536,"max_tokens":65536,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/baidu/ERNIE-4.5-VL-424B-A47B-Base-PT":{"mode":"chat","base_model":"ERNIE-4.5-VL-424B-A47B-Base-PT","provider":"huggingface","source":"https://huggingface.co/baidu/ERNIE-4.5-VL-424B-A47B-Base-PT","max_input_tokens":123000,"max_tokens":123000,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":123000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.5V":{"mode":"chat","base_model":"GLM-4.5V","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.5V","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.6V-Flash":{"mode":"chat","base_model":"GLM-4.6V-Flash","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.6V-Flash","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.7":{"mode":"chat","base_model":"GLM-4.7","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/baseten/zai-org/GLM-4.7":{"mode":"chat","base_model":"GLM-4.7","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":200000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_prompt_caching":true,"supports_reasoning":false,"supports_vision":false},"huggingface/deepinfra/zai-org/GLM-4.7":{"mode":"chat","base_model":"GLM-4.7","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/aisingapore/Gemma-SEA-LION-v4-27B-IT":{"mode":"chat","base_model":"Gemma-SEA-LION-v4-27B-IT","provider":"huggingface","source":"https://huggingface.co/aisingapore/Gemma-SEA-LION-v4-27B-IT","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/NousResearch/Hermes-3-Llama-3.1-70B":{"mode":"chat","base_model":"Hermes-3-Llama-3.1-70B","provider":"huggingface","source":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/swiss-ai/Apertus-8B-Instruct-2509":{"mode":"chat","base_model":"Apertus-8B-Instruct-2509","provider":"huggingface","source":"https://huggingface.co/swiss-ai/Apertus-8B-Instruct-2509","max_input_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"DeepSeek-R1-0528","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-R1-0528":{"mode":"chat","base_model":"DeepSeek-R1-0528","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528","max_input_tokens":163840,"max_tokens":163840,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/swiss-ai/Apertus-70B-Instruct-2509":{"mode":"chat","base_model":"Apertus-70B-Instruct-2509","provider":"huggingface","source":"https://huggingface.co/swiss-ai/Apertus-70B-Instruct-2509","max_input_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-235B-A22B-Thinking-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Thinking-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_response_schema":true},"huggingface/deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507":{"mode":"chat","base_model":"Qwen3-235B-A22B-Thinking-2507","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/publicai/aisingapore/Qwen-SEA-LION-v4-32B-IT":{"mode":"chat","base_model":"Qwen-SEA-LION-v4-32B-IT","provider":"huggingface","source":"https://huggingface.co/aisingapore/Qwen-SEA-LION-v4-32B-IT","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"DeepSeek-V3.1","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning":true},"huggingface/deepinfra/deepseek-ai/DeepSeek-V3.1":{"mode":"chat","base_model":"DeepSeek-V3.1","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/ovhcloud/Qwen/Qwen2.5-VL-72B-Instruct":{"mode":"chat","base_model":"Qwen2.5-VL-72B-Instruct","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen2.5-VL-72B-Instruct","max_input_tokens":32768,"max_tokens":32768,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V3.1-Terminus":{"mode":"chat","base_model":"DeepSeek-V3.1-Terminus","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1-Terminus","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/deepseek-ai/DeepSeek-V3.1-Terminus":{"mode":"chat","base_model":"DeepSeek-V3.1-Terminus","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1-Terminus","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-8B":{"mode":"chat","base_model":"DeepSeek-R1-Distill-Llama-8B","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/alpindale/WizardLM-2-8x22B":{"mode":"chat","base_model":"WizardLM-2-8x22B","provider":"huggingface","source":"https://huggingface.co/alpindale/WizardLM-2-8x22B","max_input_tokens":65535,"max_tokens":65535,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/zai-org/GLM-4.6":{"mode":"chat","base_model":"GLM-4.6","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.6","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-4.6":{"mode":"chat","base_model":"GLM-4.6","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4.6","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/Qwen/Qwen3-235B-A22B":{"mode":"chat","base_model":"Qwen3-235B-A22B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B","max_input_tokens":40960,"max_tokens":40960,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40960},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true},"huggingface/nscale/Qwen/Qwen3-235B-A22B":{"mode":"chat","base_model":"Qwen3-235B-A22B","provider":"huggingface","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B","max_input_tokens":32000,"max_tokens":32000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_parallel_function_calling":true,"supports_system_messages":true},"huggingface/novita/zai-org/GLM-4-32B-0414":{"mode":"chat","base_model":"GLM-4-32B-0414","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-4-32B-0414","max_input_tokens":32000,"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/deepinfra/zai-org/GLM-5.1":{"mode":"chat","base_model":"GLM-5.1","provider":"huggingface","source":"https://huggingface.co/zai-org/GLM-5.1","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M1-80k":{"mode":"chat","base_model":"MiniMax-M1-80k","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M1-80k","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/deepseek-ai/DeepSeek-V3.2-Exp":{"mode":"chat","base_model":"DeepSeek-V3.2-Exp","provider":"huggingface","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":163840},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M2":{"mode":"chat","base_model":"MiniMax-M2","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"huggingface/novita/MiniMaxAI/MiniMax-M2.1":{"mode":"chat","base_model":"MiniMax-M2.1","provider":"huggingface","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.1","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":204800},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/GLM-5.2":{"mode":"chat","base_model":"GLM-5.2","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/GLM-5.3":{"mode":"chat","base_model":"glm-5.3","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":1048576,"supports_vision":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/DeepSeek-V4-Pro":{"mode":"chat","base_model":"deepseek-v4-pro","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/Kimi-K3":{"mode":"chat","base_model":"Kimi-K3","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/Qwen3.8-27B":{"mode":"chat","base_model":"qwen3.8-27b","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":262144,"max_tokens":262144,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/GLM-5.3-Flash":{"mode":"chat","base_model":"glm-5.3-flash","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":196608,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/DeepSeek-V4.1-Flash":{"mode":"chat","base_model":"deepseek-v4.1-flash","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/DeepSeek-V4-Flash-0731-Fast":{"mode":"chat","base_model":"DeepSeek-V4-Flash-0731-Fast","provider":"wafer","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/mistral.pixtral-large-2502-v1:0":{"mode":"chat","base_model":"pixtral-large","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_reasoning":false,"provider":"bedrock","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_tool_choice":false},"bedrock_mantle/mistral.pixtral-large-2502-v1:0":{"mode":"chat","base_model":"pixtral-large","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_reasoning":false,"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_tool_choice":false},"bedrock/us.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-05-18","provider":"bedrock","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"supports_sampling_params":false},"bedrock_mantle/us.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-05-18","provider":"bedrock_mantle","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"supports_sampling_params":false},"bedrock_mantle/deepseek.r1-v1:0":{"mode":"chat","base_model":"deepseek-r1","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":32768,"max_tokens":32768,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_reasoning":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_tool_choice":false},"bedrock/jp.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":65535,"max_tokens":65535,"supported_modalities":["text","image","video","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"bedrock","supports_service_tier":true,"service_tiers":["default","flex","priority"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/jp.amazon.nova-2-lite-v1:0":{"mode":"chat","base_model":"nova-2-lite","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":65535,"max_tokens":65535,"supported_modalities":["text","image","video","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65535},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_service_tier":true},"bedrock/in.openai.gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-02-16","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"supports_pdf_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/in.openai.gpt-5.6-terra":{"mode":"chat","base_model":"gpt-5.6-terra","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-02-16","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"supports_pdf_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/us.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-04-20","provider":"bedrock","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"supports_sampling_params":false},"bedrock_mantle/us.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-04-20","provider":"bedrock_mantle","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"supports_sampling_params":false},"bedrock/ca.amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":300000,"max_output_tokens":10000,"max_tokens":10000,"supported_modalities":["text","image","video","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"bedrock","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/ca.amazon.nova-lite-v1:0":{"mode":"chat","base_model":"nova-lite","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":300000,"max_output_tokens":10000,"max_tokens":10000,"supported_modalities":["text","image","video","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"bedrock_mantle","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":10000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock/in.openai.gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-02-16","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"supports_pdf_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/in.openai.gpt-5.6-luna":{"mode":"chat","base_model":"gpt-5.6-luna","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":922000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-02-16","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":false,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-gov-east-1","us-west-1","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"supports_pdf_input":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-flash-lite-latest":{"mode":"chat","base_model":"gemini-flash-lite","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supported_modalities":["text","image","video","audio","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_reasoning":true,"reasoning_effort_levels":["minimal","low","medium","high"],"provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_audio_output":false,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_web_search":true,"supports_native_streaming":true},"vertex_ai/gemini-flash-latest":{"mode":"chat","base_model":"gemini-flash","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supported_modalities":["text","image","video","audio","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_audio_input":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["minimal","low","medium","high"],"provider":"vertex_ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":65536},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_audio_output":false,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_web_search":true,"supports_native_streaming":true},"gpt-5.3-codex-spark":{"mode":"chat","base_model":"gpt-5.3-codex-spark","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":100000,"max_output_tokens":32000,"max_tokens":32000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh"],"knowledge_cutoff":"2025-08-31","provider":"openai","web_search_billing_unit":"per_request","model_parameters":[{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions"]}]},"cloudflare/workers-ai/@cf/qwen/qwen3-30b-a3b-fp8":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/qwen/qwen3-30b-a3b-fp8":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/qwen/qwen3.8-27b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","xhigh"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/qwen/qwen3.8-27b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","xhigh"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/qwen/qwq-32b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":24000,"max_output_tokens":24000,"max_tokens":24000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":true,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/qwen/qwq-32b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":24000,"max_output_tokens":24000,"max_tokens":24000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":true,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/qwen/qwen2.5-coder-32b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/qwen/qwen2.5-coder-32b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32768,"max_output_tokens":32768,"max_tokens":32768,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32768},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/deepseek-ai/deepseek-v4-pro-0813":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/deepseek-ai/deepseek-v4-pro-0813":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/deepseek-ai/deepseek-v4-flash-0731":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/deepseek-ai/deepseek-v4-flash-0731":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/deepseek-ai/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":80000,"max_output_tokens":80000,"max_tokens":80000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":true,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/deepseek-ai/deepseek-r1-distill-qwen-32b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":80000,"max_output_tokens":80000,"max_tokens":80000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":true,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/mistralai/mistral-small-3.1-24b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/mistralai/mistral-small-3.1-24b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/nvidia/nemotron-3-120b-a12b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/nvidia/nemotron-3-120b-a12b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":256000,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":256000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"cloudflare","reasoning_effort_levels":["none","high"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/google/gemma-4-26b-a4b-it":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":256000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"workers-ai","reasoning_effort_levels":["none","high"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/zai-org/glm-5.2":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","high","max"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/zai-org/glm-5.2":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","high","max"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/zai-org/glm-5.3-flash":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/zai-org/glm-5.3-flash":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/zai-org/glm-5.3":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/zai-org/glm-5.3":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1310720,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/zai-org/glm-4.7-flash":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/zai-org/glm-4.7-flash":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-guard-3-8b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-guard-3-8b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131072,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-3.1-8b-instruct-fp8":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-3.1-8b-instruct-fp8":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":32000,"max_output_tokens":32000,"max_tokens":32000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-3.2-3b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":80000,"max_output_tokens":80000,"max_tokens":80000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-3.2-3b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":80000,"max_output_tokens":80000,"max_tokens":80000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":80000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-3.2-1b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":60000,"max_output_tokens":60000,"max_tokens":60000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":60000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-3.2-1b-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":60000,"max_output_tokens":60000,"max_tokens":60000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":60000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-4-scout-17b-16e-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-3.3-70b-instruct-fp8-fast":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":24000,"max_output_tokens":24000,"max_tokens":24000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-3.3-70b-instruct-fp8-fast":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":24000,"max_output_tokens":24000,"max_tokens":24000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":24000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/meta/llama-3.2-11b-vision-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/meta/llama-3.2-11b-vision-instruct":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":false,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/ibm-granite/granite-4.0-h-micro":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/ibm-granite/granite-4.0-h-micro":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":131000,"max_output_tokens":131000,"max_tokens":131000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":false,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/openai/gpt-oss-20b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/openai/gpt-oss-20b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/openai/gpt-oss-120b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/openai/gpt-oss-120b":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":16384,"max_tokens":16384,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16384},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/moonshotai/kimi-k2.6":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","high"],"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/moonshotai/kimi-k2.6":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":256000,"max_tokens":256000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","high"],"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":256000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"cloudflare/workers-ai/@cf/moonshotai/kimi-k2.7-code":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"cloudflare","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"workers-ai/workers-ai/@cf/moonshotai/kimi-k2.7-code":{"mode":"chat","base_model":"cloudflare/workers-ai/","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"provider":"workers-ai","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/databricks-kimi-k2-7-code":{"mode":"chat","base_model":"databricks/databricks-kimi-k2-7-code","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"databricks","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/system.ai.kimi-k2-7-code":{"mode":"chat","base_model":"databricks/databricks-kimi-k2-7-code","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"databricks","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/kimi-k2-7-code":{"mode":"chat","base_model":"databricks/databricks-kimi-k2-7-code","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_vision":true,"supports_video_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"provider":"databricks","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_system_messages":true,"supports_web_search":true},"bedrock_mantle/mistral.ministral-3-8b-instruct":{"mode":"chat","base_model":"ministral-3-8b-instruct","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.ministral-3-14b-instruct":{"mode":"chat","base_model":"ministral-3-14b-instruct","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/zai.glm5":{"mode":"chat","base_model":"glm5","provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/google.gemma-3-12b-it":{"mode":"chat","base_model":"gemma-3-12b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/writer.palmyra-vision-7b":{"mode":"chat","base_model":"palmyra-vision-7b","max_input_tokens":4000,"max_output_tokens":4000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_function_calling":true,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","ap-northeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","sa-east-1"],"max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/moonshotai.kimi-k3":{"mode":"chat","base_model":"kimi-k3","provider":"bedrock_mantle","supports_prompt_caching":true,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supports_function_calling":true,"supports_tool_choice":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_input_tokens":1000000,"max_tokens":1000000,"source":"https://aws.amazon.com/bedrock/pricing/","min_output_tokens":16,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1000000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.ministral-3-3b-instruct":{"mode":"chat","base_model":"ministral-3-3b-instruct","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.mistral-large-3-675b-instruct":{"mode":"chat","base_model":"mistral-large-3-675b-instruct","max_input_tokens":256000,"max_output_tokens":32000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1"],"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/deepseek.v3.2":{"mode":"chat","base_model":"v3.2","max_input_tokens":164000,"max_output_tokens":8000,"supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-north-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.voxtral-small-24b-2507":{"mode":"chat","base_model":"voxtral-small-24b-2507","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_vision":false,"supports_audio_output":false,"supports_function_calling":true,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-next-80b-a3b-instruct":{"mode":"chat","base_model":"qwen3-next-80b-a3b-instruct","max_input_tokens":256000,"max_output_tokens":8000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.devstral-2-123b":{"mode":"chat","base_model":"devstral-2-123b","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":256000,"max_output_tokens":32000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4"],"is_deprecated":true,"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/minimax.minimax-m2.1":{"mode":"chat","base_model":"minimax-m2.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":196000,"max_output_tokens":8000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-coder-next":{"mode":"chat","base_model":"qwen3-coder-next","max_input_tokens":256000,"max_output_tokens":16000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","eu-west-2","ap-southeast-2"],"max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/deepseek.v3.1":{"mode":"chat","base_model":"v3.1","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-2","us-west-2","eu-north-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","us-east-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-235b-a22b-2507":{"mode":"chat","base_model":"qwen3-235b-a22b-2507","max_input_tokens":256000,"max_output_tokens":8000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","eu-west-1","sa-east-1","us-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-32b":{"mode":"chat","base_model":"qwen3-32b","max_input_tokens":32000,"max_output_tokens":8000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/zai.glm-4.7":{"mode":"chat","base_model":"glm-4.7","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"zai.glm-5","max_input_tokens":203000,"max_output_tokens":4000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-north-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4"],"is_deprecated":true,"max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-vl-235b-a22b-instruct":{"mode":"chat","base_model":"qwen3-vl-235b-a22b-instruct","max_input_tokens":256000,"max_output_tokens":8000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-coder-30b-a3b-instruct":{"mode":"chat","base_model":"qwen3-coder-30b-a3b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":256000,"max_output_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1"],"is_deprecated":true,"max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/google.gemma-3-27b-it":{"mode":"chat","base_model":"gemma-3-27b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.voxtral-mini-3b-2507":{"mode":"chat","base_model":"voxtral-mini-3b-2507","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32000,"max_output_tokens":8192,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_vision":false,"supports_audio_output":false,"supports_function_calling":true,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":32000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":32000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/minimax.minimax-m2":{"mode":"chat","base_model":"minimax-m2","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":1000000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/google.gemma-3-4b-it":{"mode":"chat","base_model":"gemma-3-4b-it","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":128000,"max_output_tokens":8000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_system_messages":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"is_deprecated":true,"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/mistral.magistral-small-2509":{"mode":"chat","base_model":"magistral-small-2509","max_input_tokens":128000,"max_output_tokens":40000,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supports_tool_choice":true,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","sa-east-1","ap-southeast-3","ap-southeast-4","eu-central-1","eu-north-1"],"max_tokens":40000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":40000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/zai.glm-4.7-flash":{"mode":"chat","base_model":"glm-4.7-flash","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":203000,"max_output_tokens":4000,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4"],"is_deprecated":true,"max_tokens":4000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4000,"range":{"min":1,"max":4000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/minimax.minimax-m2.5":{"mode":"chat","base_model":"minimax-m2.5","max_input_tokens":196000,"max_output_tokens":8000,"supports_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-2","eu-central-1","eu-north-1","eu-south-1","eu-west-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","ap-southeast-4"],"max_tokens":8000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/qwen.qwen3-coder-480b-a35b-instruct":{"mode":"chat","base_model":"qwen3-coder-480b-a35b-instruct","deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html","max_input_tokens":128000,"max_output_tokens":16000,"supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-2","us-west-2","eu-north-1","eu-west-2","ap-northeast-1","ap-south-1","ap-southeast-2","ap-southeast-3","sa-east-1","us-east-1"],"is_deprecated":true,"max_tokens":16000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":16000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/GLM-5.1":{"mode":"chat","base_model":"GLM-5.1","provider":"wafer","max_tokens":202752,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":202752},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/Kimi-K2.6":{"mode":"chat","base_model":"Kimi-K2.6","provider":"wafer","max_tokens":262144,"supports_function_calling":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/MiniMax-M3":{"mode":"chat","base_model":"MiniMax-M3","provider":"wafer","max_tokens":1048576,"supports_function_calling":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/Qwen3.5-397B-A17B":{"mode":"chat","base_model":"Qwen3.5-397B-A17B","provider":"wafer","max_tokens":262133,"supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":262133},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/glm5.2-fast":{"mode":"chat","base_model":"glm5.2-fast","provider":"wafer","max_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"wafer/kimi-k3-fast":{"mode":"chat","base_model":"kimi-k3-fast","provider":"wafer","max_tokens":1048576,"supports_function_calling":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-5.6-sol":{"mode":"responses","base_model":"gpt-5.6-sol","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"bedrock_mantle/us-gov-west-1/openai.gpt-5.5":{"mode":"responses","base_model":"gpt-5.5","max_input_tokens":272000,"max_output_tokens":128000,"max_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock_mantle","model_parameters":[{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/responses"]}]},"runware/zai:glm@5.3-flash":{"mode":"chat","base_model":"glm-5.3-flash","provider":"runware","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"runware/google:gemini@3.7-flash":{"mode":"chat","base_model":"gemini-3.7-flash","provider":"runware","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"runware/deepseek:v4@flash":{"mode":"chat","base_model":"deepseek-v4-flash","provider":"runware","model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","range":{"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation. In-geo endpoint is 10% higher."},"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_anthropic_thinking_payload":true,"supports_tool_choice":true,"prompt_cache_min_tokens":4096,"provider":"databricks","supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"databricks/claude-sonnet-4-5":{"mode":"chat","base_model":"claude-sonnet-4-5","provider":"databricks","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":200000,"max_tokens":64000,"max_output_tokens":64000,"supports_vision":true,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":64000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"google/gemma-4-26b-a4b-it-maas":{"mode":"chat","base_model":"gemma-4","max_input_tokens":262144,"max_output_tokens":8192,"max_tokens":8192,"source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models","supported_regions":["global"],"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","tpm":250000,"rpm":10,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":8192},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"gemini-3.1-flash-lite-image":{"mode":"image_generation","base_model":"gemini-3.1-flash-lite-image","max_input_tokens":65536,"max_output_tokens":66000,"source":"https://ai.google.dev/gemini-api/docs/pricing#gemini-3.1-flash-lite-image","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":true,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"provider":"gemini","max_tokens":66000,"rpm":null,"supports_web_search":false,"tpm":null,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":66000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_decisions":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6-luna-2026-09-22":{"mode":"chat","base_model":"gpt-6-luna","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_decisions":true,"supports_image_input":true,"supports_native_structured_output":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6-sol-2026-09-22":{"mode":"chat","base_model":"gpt-6-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_image_input":true,"supports_native_structured_output":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/gpt-6.1-sol-2026-09-29":{"mode":"chat","base_model":"gpt-6.1-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/openai.gpt-6-sol":{"mode":"responses","base_model":"gpt-6-sol","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supports_web_search":true,"provider":"bedrock_mantle","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/responses"],"provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/responses"],"provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_reasoning_with_tool_calls":false,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6.1-sol-pro":{"mode":"chat","base_model":"gpt-6.1-sol-pro","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6.1-sol-pro:batch":{"mode":"chat","base_model":"gpt-6.1-sol-pro:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"openrouter/openai/gpt-6.1-sol:batch":{"mode":"chat","base_model":"gpt-6.1-sol:batch","provider":"openrouter","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true,"source":"https://openrouter.ai/api/v1/models","supports_audio_input":false,"supports_web_search":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":true},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6.1-sol","supported_endpoints":["/v1/responses","/v1/chat/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","thinking_always_on":true,"supports_low_reasoning_effort":true,"supports_sampling_params":false,"supports_reasoning_with_tool_calls":false,"supports_native_structured_output":true,"supports_image_input":true,"supports_audio_input":false,"supports_audio_output":false,"supports_video_input":false,"supports_file_search":true,"supports_image_gen_tool":true,"supports_code_execution":true,"supports_service_tier":true,"regional_endpoint_uplift_multiplier":1.1,"supports_reasoning_disable":false,"supports_async_tools":true,"supports_custom_tools":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"service_tiers":["auto","default","flex","priority","fast","ultrafast"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.openai.gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-sol.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_bedrock_runtime_chat_completions_response_format":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_sampling_params":false,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","mx-central-1","af-south-1","sa-east-1"],"max_tokens":128000,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_converse_tool_result_images":false},"openai.gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-sol.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock_mantle/openai.gpt-6.1-sol":{"mode":"responses","base_model":"gpt-6.1-sol","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-sol.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1"],"max_tokens":128000,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"supports_service_tier":true,"service_tiers":["auto","default","ultrafast"],"bedrock_mantle_base_path":"openai/v1","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.openai.gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-sol.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_bedrock_runtime_chat_completions_response_format":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_sampling_params":false,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1"],"max_tokens":128000,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_converse_tool_result_images":false},"opencode-zen/gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","max_input_tokens":1050000,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_prompt_caching":true,"supports_native_streaming":true,"supports_system_messages":true,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"opencode-zen/gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","max_input_tokens":1050000,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_prompt_caching":true,"supports_native_streaming":true,"supports_system_messages":true,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"opencode-zen/gpt-6.1-sol":{"mode":"chat","base_model":"gpt-6.1-sol","provider":"opencode-zen","source":"https://opencode.ai/docs/zen/","max_input_tokens":1050000,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_reasoning_disable":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_prompt_caching":true,"supports_native_streaming":true,"supports_system_messages":true,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-sol-2026-09-22":{"mode":"chat","base_model":"gpt-6-sol","provider":"azure","max_input_tokens":922000,"max_output_tokens":128000,"supports_vision":true,"supports_image_input":true,"supported_output_modalities":["text"],"supports_reasoning":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_none_reasoning_effort":true,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_reasoning_with_tool_calls":false,"supports_prompt_caching":true,"supports_system_messages":true,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/gpt-6-sol-2026-09-22":{"mode":"chat","base_model":"gpt-6-sol","provider":"azure","max_input_tokens":922000,"max_output_tokens":128000,"supports_vision":true,"supports_image_input":true,"supported_output_modalities":["text"],"supports_reasoning":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_none_reasoning_effort":true,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_reasoning_with_tool_calls":false,"supports_prompt_caching":true,"supports_system_messages":true,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/us/gpt-6-luna-2026-09-22":{"mode":"chat","base_model":"gpt-6-luna","provider":"azure","max_input_tokens":922000,"max_output_tokens":128000,"supports_vision":true,"supports_image_input":true,"supported_output_modalities":["text"],"supports_reasoning":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_none_reasoning_effort":true,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_reasoning_with_tool_calls":false,"supports_prompt_caching":true,"supports_system_messages":true,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/eu/gpt-6-luna-2026-09-22":{"mode":"chat","base_model":"gpt-6-luna","provider":"azure","max_input_tokens":922000,"max_output_tokens":128000,"supports_vision":true,"supports_image_input":true,"supported_output_modalities":["text"],"supports_reasoning":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_none_reasoning_effort":true,"source":"https://azure.microsoft.com/en-us/blog/gpt-6-astra-sol-and-luna-for-production-agents-in-microsoft-foundry/","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_tool_choice":true,"supports_reasoning_with_tool_calls":false,"supports_prompt_caching":true,"supports_system_messages":true,"max_tokens":128000,"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"supports_service_tier":true,"service_tiers":["auto","default","flex","priority"],"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_parallel_function_calling":true,"supports_forced_tool_use":false,"thinking_always_on":true,"supports_forced_tool_choice":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_web_search":true,"provider":"anthropic","web_search_billing_unit":"per_request","supports_thinking_cache_preservation":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":true,"supports_reasoning_disable":false,"supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"stop_sequences","label":"Stop Sequence","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","minElements":1,"maxElements":4}},{"id":"thinking","label":"Thinking","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}]},{"id":"output_config","label":"output_config","type":"select","accesorKey":"effort","default":{"type":"text"},"options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","type":"select","default":"high","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","schema":{},"strict":true}}]},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_parallel_function_calling":true},"claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"anthropic","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"Enable web search tool for real-time information retrieval.","id":"web_search","label":"Web Search","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"bedrock","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_bedrock_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false,"supports_bedrock_native_structured_output_by_default":true},"eu.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"bedrock","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_bedrock_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false,"supports_bedrock_native_structured_output_by_default":true},"au.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"bedrock","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_bedrock_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false,"supports_bedrock_native_structured_output_by_default":true},"jp.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"bedrock","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_bedrock_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false,"supports_bedrock_native_structured_output_by_default":true},"global.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/overview","supports_web_search":true,"provider":"bedrock","web_search_billing_unit":"per_request","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","supports_image_input":true,"supports_system_messages":true,"supports_native_streaming":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_tool_search":true,"supports_bedrock_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_audio_input":false,"supports_audio_output":false,"supports_bedrock_native_structured_output_by_default":true},"FW-GLM-5.3":{"mode":"chat","base_model":"fw-glm-5.3","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","supports_reasoning_summary_detailed":false,"supports_reasoning_summary_concise":false,"supports_reasoning_summary_auto":true,"supported_endpoints":["/v1/chat/completions"],"model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","default":1,"range":{"min":0,"max":2,"step":0.01},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":131072},"supported_endpoints":["/v1/chat/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":5,"minElements":1},"supported_endpoints":["/v1/chat/completions"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","default":1,"range":{"min":0,"max":1,"step":0.01},"supported_endpoints":["/v1/chat/completions"]},{"id":"frequency_penalty","label":"Frequency Penalty","helpText":"Number between 0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions"]},{"id":"presence_penalty","label":"Presence Penalty","helpText":"Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":0,"supported_endpoints":["/v1/chat/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","default":1,"range":{"min":1,"max":10,"step":1},"supported_endpoints":["/v1/chat/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Select the level of reasoning effort. Higher levels may use more tokens and provide more thorough reasoning in the model's response.","type":"select","default":"medium","options":[{"value":"low","label":"Low"},{"value":"medium","label":"Medium"},{"value":"high","label":"High"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Structured output","value":"json_schema","subFields":[{"id":"json_schema","label":"JSON Schema","helpText":"A JSON schema that the model will follow when generating the response.","type":"json","default":{"name":"","strict":true,"schema":{}}}]},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions"]}],"supports_parallel_function_calling":true,"reasoning_effort_levels":["low","medium","high"],"supports_none_reasoning_effort":false,"supports_pdf_input":false},"gemini-4-argon":{"mode":"chat","base_model":"gemini-4-argon","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"supports_code_execution":true,"supports_computer_use":true,"supports_service_tier":true,"provider":"gemini","service_tiers":["default","flex","priority"],"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/responses"]},{"array":{"maxElements":5,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_continuation":true},"gemini/gemini-4-argon":{"mode":"chat","base_model":"gemini-4-argon","max_input_tokens":1048576,"max_output_tokens":1048576,"prompt_cache_min_tokens":4096,"rpm":2000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"web_search_billing_unit":"per_query","provider":"gemini","supports_code_execution":true,"supports_service_tier":true,"service_tiers":["priority","flex"],"max_tokens":1048576,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"vertex_ai/gemini-4-argon":{"mode":"chat","base_model":"gemini-4-argon","max_input_tokens":1048576,"max_output_tokens":1048576,"prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","supports_audio_output":false,"supports_code_execution":true,"supports_service_tier":true,"service_tiers":["priority"],"max_tokens":1048576,"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"azure/claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","supports_mid_conversation_system":true,"deprecation_date":"2027-06-30","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":false,"supports_web_search":true,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://management.azure.com/subscriptions/c873328e-b572-4770-8dff-aaeb6f1f0e79/providers/Microsoft.CognitiveServices/locations/eastus2/models?api-version=2024-10-01","provider":"azure","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"server_tools":{"web_search":"web_search_20250305"}},"apac.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_parallel_function_calling":true,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"eu.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"jp.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"au.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_mid_conversation_system_messages":false,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-mythos-5-1":{"mode":"chat","base_model":"claude-mythos-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-opus-4-6":{"mode":"chat","base_model":"claude-opus-4-6","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"tool_use_system_prompt_tokens":346,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"source":"https://aws.amazon.com/bedrock/pricing/","supports_web_search":true,"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"supports_reasoning_disable":false,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_safeguards":true,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-mythos-preview":{"mode":"chat","base_model":"claude-mythos","max_input_tokens":1000000,"max_output_tokens":128000,"thinking_always_on":true,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_output_config":true,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-opus-4-7":{"mode":"chat","base_model":"claude-opus-4-7","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":2048,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"ap-northeast-1"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-opus-4-8":{"mode":"chat","base_model":"claude-opus-4-8","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"ap-northeast-1"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-sonnet-4-6":{"mode":"chat","base_model":"claude-sonnet-4-6","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":64000,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_max_reasoning_effort":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-sonnet-4-6","litellm_provider":"bedrock"},"provider":"bedrock","max_tokens":64000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"apac.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":1024,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"au.anthropic.claude-fable-5":{"mode":"chat","base_model":"claude-fable-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"au.anthropic.claude-mythos-preview":{"mode":"chat","base_model":"claude-mythos","max_input_tokens":1000000,"max_output_tokens":128000,"thinking_always_on":true,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_output_config":true,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/anthropic.claude-haiku-4-5":{"mode":"chat","base_model":"claude-haiku-4-5","supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"provider":"bedrock","supports_web_search":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_native_streaming":true,"supports_reasoning_disable":true,"default_reasoning_effort":"medium","reasoning_effort_levels":["low","medium","high","xhigh","max"],"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/au.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","token_count_model":{"model":"au.anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"ap-southeast-4"},"bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/au.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"supports_native_effort":true,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/eu.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/eu.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/global.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"token_count_model":{"model":"global.anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/global.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"supports_forced_tool_choice":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/in.anthropic.claude-haiku-4-5-20251001-v1:0":{"mode":"chat","base_model":"claude-haiku-4-5","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"knowledge_cutoff":"2025-02-28","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":1,"helpText":"What sampling temperature to use, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","id":"temperature","label":"Temperature","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":64000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"default":1,"helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","id":"top_p","label":"Top P","range":{"max":1,"min":0,"step":0.01},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"default":50,"helpText":"The top_k parameter is used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence.","id":"top_k","label":"Top K","range":{"max":100,"min":1},"type":"number"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"enabled","value":"enabled"}],"type":"select"},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/in.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_mid_conversation_system_messages":false},"bedrock/in.anthropic.claude-sonnet-5":{"mode":"chat","base_model":"claude-sonnet-5","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-01-31","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_native_structured_output":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/jp.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/jp.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-east-1/anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-haiku-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov-west-1/anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-sonnet-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us-gov.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"token_count_model":{"model":"us.anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output_by_default":true,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/us.anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"},{"label":"between_tools","value":"between_tools"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.anthropic.claude-mythos-5-1":{"mode":"chat","base_model":"claude-mythos-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us-gov.anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-haiku-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock","supports_safeguards":true,"supports_native_effort":true,"supports_mid_conversation_system_messages":false,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"disabled","value":"disabled"},{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-mythos-5":{"mode":"chat","base_model":"claude-mythos-5","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-mythos-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-mythos-5-1":{"mode":"chat","base_model":"claude-mythos-5-1","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"us.anthropic.claude-mythos-preview":{"mode":"chat","base_model":"claude-mythos","max_input_tokens":1000000,"max_output_tokens":128000,"thinking_always_on":true,"supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_output_config":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","supports_safeguards":true,"max_tokens":128000,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":16384,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"max":128000,"min":1},"supported_endpoints":["/v1/chat/completions","/v1/completions"],"type":"number"},{"array":{"maxElements":4,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop_sequences","label":"Stop Sequence","type":"array"},{"accesorKey":"type","default":{"type":"text"},"id":"thinking","label":"Thinking","options":[{"label":"adaptive","value":"adaptive"}],"type":"select"},{"accesorKey":"effort","default":{"type":"text"},"id":"output_config","label":"output_config","options":[{"label":"max","value":"max"},{"label":"xhigh","value":"xhigh"},{"label":"high","value":"high"},{"label":"medium","value":"medium"},{"label":"low","value":"low"}],"type":"select"},{"default":"medium","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Lower values favour speed and economical token usage; higher values favour more complete reasoning.","id":"reasoning_effort","label":"Reasoning Effort","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"metadata","label":"Metadata","helpText":"An external identifier for the user who is associated with the request.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"gpt-6.1-terra":{"mode":"chat","base_model":"gpt-6.1-terra","max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6.1-terra","supported_endpoints":["/v1/responses","/v1/chat/completions","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","web_search_billing_unit":"per_request","supports_custom_tools":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_reasoning_with_tool_calls":false,"default_reasoning_effort":"medium","thinking_always_on":true,"supports_low_reasoning_effort":true,"supports_sampling_params":false,"supports_native_structured_output":true,"supports_image_input":true,"supports_audio_input":false,"supports_audio_output":false,"supports_video_input":false,"supports_file_search":true,"supports_image_gen_tool":true,"supports_code_execution":true,"supports_service_tier":true,"regional_endpoint_uplift_multiplier":1.1,"supports_reasoning_disable":false,"supports_async_tools":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"service_tiers":["auto","default","flex","priority","fast"],"max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"n","label":"Number of Responses","helpText":"How many chat completion choices to generate for each input message.","type":"number","range":{"min":1,"max":10,"step":1},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"none","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"},{"label":"Concise","value":"concise"},{"label":"Detailed","value":"detailed"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":1050000,"max_output_tokens":128000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/responses"],"provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":true,"supports_low_reasoning_effort":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"max_tokens":128000,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_image_input":true,"supports_prompt_cache_breakpoint":true,"supports_native_structured_output":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true},"claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"provider_specific_entry":{"us":1.1},"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"supports_anthropic_compaction":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_fast_mode":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_speed":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"anthropic","web_search_billing_unit":"per_request","max_tokens":1000000,"inference_geo_us_multiplier":1.1,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"supports_reasoning_with_tool_calls":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":true,"supports_safeguards":true,"supports_service_tier":true,"supports_native_effort":true,"supports_reasoning_effort":true,"supports_dynamic_reasoning_budget":false,"supports_response_schema_with_tools":true,"service_tiers":["auto","standard_only","fast"]},"anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_mid_conversation_system_messages":false},"global.anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output_by_default":true,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"global.anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_mid_conversation_system_messages":false},"us.anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output_by_default":true,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supports_mid_conversation_system_messages":false},"eu.anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output_by_default":true,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"supports_audio_input":false,"supports_audio_output":false,"token_count_model":{"model":"anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"eu-west-1"},"supports_mid_conversation_system_messages":false},"bedrock_mantle/anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"prompt_cache_min_tokens":512,"tool_use_system_prompt_tokens":346,"supports_reasoning":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_native_effort":true,"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_output_config":true,"supports_sampling_params":false,"supports_assistant_prefill":false,"supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_parallel_tool_use_config":true,"supports_computer_use":true,"supports_mid_conversation_system":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_vision":true,"supports_image_input":true,"supports_web_search":true,"supports_safeguards":true,"supports_native_streaming":true,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"token_count_model":{"model":"anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-west-1/anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","provider":"bedrock_mantle","source":"https://aws.amazon.com/bedrock/pricing/","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"prompt_cache_min_tokens":512,"tool_use_system_prompt_tokens":346,"supports_reasoning":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_native_effort":true,"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_output_config":true,"supports_sampling_params":false,"supports_assistant_prefill":false,"supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_tool_use_config":true,"supports_computer_use":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_vision":true,"supports_safeguards":true,"supports_audio_input":false,"supports_audio_output":false,"token_count_model":{"model":"anthropic.claude-fable-5-2","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"supported_regions":["us-gov-west-1","us-gov-east-1"],"supports_mid_conversation_system_messages":true},"vertex_ai/claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-11-09","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","default_reasoning_effort":"high","max_tokens":128000,"provider_specific_entry":{"us":1.1},"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_fast_mode":true,"supports_forced_tool_choice":false,"supports_output_config":true,"supports_reasoning_disable":false,"supports_safeguards":true,"supports_speed":true,"supports_web_search":true,"tool_use_system_prompt_tokens":346,"supports_mid_conversation_system_messages":false},"vertex_ai/claude-fable-5-2@default":{"mode":"chat","base_model":"claude-fable-5-2","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-11-09","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","default_reasoning_effort":"high","max_tokens":128000,"provider_specific_entry":{"us":1.1},"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_forced_tool_choice":false,"supports_output_config":true,"supports_reasoning_disable":false,"supports_safeguards":true,"supports_speed":true,"supports_web_search":true,"tool_use_system_prompt_tokens":346,"supports_mid_conversation_system_messages":false},"azure/claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"thinking_always_on":true,"provider":"azure","max_tokens":128000,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_safeguards":true,"supports_mid_conversation_system_messages":false},"gemini-4-carbon":{"mode":"chat","base_model":"gemini-4-carbon","prompt_cache_min_tokens":4096,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"supports_code_execution":true,"supports_computer_use":true,"supports_service_tier":true,"provider":"gemini","service_tiers":["default","flex","priority"],"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"helpText":"A list of tools the model may call.","id":"promptTools","label":"Prompt Tools (Functions)","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":"medium","helpText":"Set the thinking level","id":"thinking_level","label":"Thinking Level","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"type":"select"},{"default":256,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":1,"max":1048576},"supported_endpoints":["/v1/responses"]},{"array":{"maxElements":5,"minElements":1,"type":"text"},"helpText":"Custom text sequences that will cause the model to stop generating.","id":"stop","label":"Stop","type":"array","supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"accesorKey":"type","default":{"type":"text"},"helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","id":"response_format","label":"Response Format","options":[{"label":"Structured output","subFields":[{"default":{"name":"","schema":{},"strict":true},"helpText":"A JSON schema that the model will follow when generating the response.","id":"json_schema","label":"JSON Schema","type":"json"}],"value":"json_schema"},{"label":"JSON","value":"json_object"},{"label":"Text","value":"text"}],"type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"default":false,"helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","id":"stream","label":"Stream","type":"boolean","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"source":"https://aistudio.google.com/docs/pricing#gemini-4-carbon","supports_reasoning_budget":false,"supports_dynamic_reasoning_budget":false,"supports_continuation":true},"gemini/gemini-4-carbon":{"mode":"chat","base_model":"gemini-4-carbon","max_input_tokens":1048576,"max_output_tokens":1048576,"prompt_cache_min_tokens":4096,"rpm":2000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://aistudio.google.com/docs/pricing#gemini-4-carbon","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"tpm":800000,"web_search_billing_unit":"per_query","provider":"gemini","max_tokens":1048576,"supports_code_execution":true,"supports_service_tier":true,"service_tiers":["priority","flex"],"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning_budget":false,"supports_dynamic_reasoning_budget":false},"vertex_ai/gemini-4-carbon":{"mode":"chat","base_model":"gemini-4-carbon","max_input_tokens":1048576,"max_output_tokens":1048576,"prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","max_tokens":1048576,"supports_audio_output":false,"supports_code_execution":true,"supports_service_tier":true,"service_tiers":["priority"],"supports_reasoning_with_tool_calls":true,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high"],"supports_reasoning_disable":false,"supports_none_reasoning_effort":false,"supports_assistant_prefill":false,"supports_response_schema_with_tools":true,"supports_sampling_params":false,"model_parameters":[{"default":4096,"helpText":"The maximum number of tokens that can be generated in the Result.","id":"max_tokens","label":"Max Tokens","range":{"min":1,"max":1048576},"type":"number","supported_endpoints":["/v1/chat/completions","/v1/completions"]}],"supports_reasoning_budget":false,"supports_dynamic_reasoning_budget":false},"openai.gpt-6.1-terra":{"mode":"chat","base_model":"gpt-6.1-terra","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-terra.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_response_schema":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_converse_tool_result_images":false,"supports_prompt_cache_breakpoint":true,"supports_web_search":true,"supports_audio_input":false,"supports_audio_output":false},"us.openai.gpt-6.1-terra":{"mode":"chat","base_model":"gpt-6.1-terra","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-terra.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_bedrock_converse_tool_result_images":false,"supports_bedrock_runtime_chat_completions_response_format":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1"],"max_tokens":128000,"supports_response_schema":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_converse_tool_result_images":false},"global.openai.gpt-6.1-terra":{"mode":"chat","base_model":"gpt-6.1-terra","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-terra.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_bedrock_converse_tool_result_images":false,"supports_bedrock_runtime_chat_completions_response_format":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","mx-central-1","af-south-1","sa-east-1"],"max_tokens":128000,"supports_response_schema":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":true,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"supports_service_tier":true,"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}],"supports_converse_tool_result_images":false},"bedrock_mantle/openai.gpt-6.1-terra":{"mode":"responses","base_model":"gpt-6.1-terra","max_input_tokens":1050000,"max_output_tokens":128000,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-terra.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1"],"max_tokens":128000,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_reasoning_disable":false,"supports_low_reasoning_effort":true,"thinking_always_on":true,"always_reasoning":true,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_parallel_function_calling":true,"supports_reasoning_with_tool_calls":false,"supports_native_streaming":true,"supports_system_messages":true,"min_output_tokens":16,"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"supports_service_tier":true,"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"azure/kimi-k3.1-pro":{"mode":"chat","base_model":"kimi-k3.1-pro","max_input_tokens":1048576,"max_output_tokens":262144,"max_tokens":262144,"reasoning_effort_levels":["low","high","max"],"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/moonshot-ai%E2%80%99s-kimi-k3-1-pro-comes-to-microsoft-foundry/4570292","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","model_parameters":[{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions"]},{"id":"max_completion_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result, including reasoning tokens.","type":"number","default":4096,"range":{"min":1,"max":262144},"supported_endpoints":["/v1/chat/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"Reasoning effort gives the model guidance on how many reasoning tokens to generate before answering. Kimi K3.1 Pro always reasons; lower values favour speed and economical token usage, higher values favour more complete reasoning.","type":"select","default":"max","options":[{"label":"Low","value":"low"},{"label":"High","value":"high"},{"label":"Max","value":"max"}],"supported_endpoints":["/v1/chat/completions"]},{"id":"stop","label":"Stop","helpText":"Custom text sequences that will cause the model to stop generating.","type":"array","array":{"type":"text","maxElements":4,"minElements":1},"supported_endpoints":["/v1/chat/completions"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions"]}]},"azure/claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","supports_mid_conversation_system":true,"deprecation_date":"2027-09-29","max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":true,"supports_web_search":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/models/haiku-5-5/migration-guide","provider":"azure","supports_native_effort":true,"supports_safeguards":true,"supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-fable-5":{"mode":"chat","base_model":"claude-platform/claude-fable-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-fable-5-1":{"mode":"chat","base_model":"claude-platform/claude-fable-5-1","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-fable-5-2":{"mode":"chat","base_model":"claude-platform/claude-fable-5-2","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"provider_specific_entry":{"us":1.1},"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","supports_adaptive_thinking":true,"supports_anthropic_compaction":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-haiku-5-5":{"mode":"chat","base_model":"claude-platform/claude-haiku-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","supports_web_search":true,"provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-opus-5":{"mode":"chat","base_model":"claude-platform/claude-opus-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-opus-5-5":{"mode":"chat","base_model":"claude-platform/claude-opus-5-5","supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/claude_platform/claude-sonnet-5-5":{"mode":"chat","base_model":"claude-platform/claude-sonnet-5-5","supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","supports_web_search":true,"provider":"bedrock","supports_mid_conversation_system_messages":false},"bedrock/us-gov-east-1/anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"supports_mid_conversation_system_messages":false},"bedrock/us-gov-west-1/anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"supports_mid_conversation_system_messages":false},"bedrock_mantle/anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-gov-west-1"],"max_tokens":128000,"supports_native_effort":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_native_streaming":true,"supports_safeguards":true,"supports_reasoning_disable":true,"default_reasoning_effort":"medium","reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_mid_conversation_system_messages":true},"bedrock_mantle/anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":true},"bedrock_mantle/in.anthropic.claude-opus-5":{"mode":"chat","base_model":"claude-opus-5","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-west-1/anthropic.claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-haiku-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_native_effort":true,"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-west-1/anthropic.claude-opus-5-5":{"mode":"chat","base_model":"claude-opus-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"supports_forced_tool_use":false,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_mid_conversation_system_messages":true},"bedrock_mantle/us-gov-west-1/anthropic.claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"source":"https://aws.amazon.com/bedrock/pricing/","token_count_model":{"model":"anthropic.claude-sonnet-5-5","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000,"supports_safeguards":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":true},"us-gov.anthropic.claude-fable-5-2":{"mode":"chat","base_model":"claude-fable-5-2","bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","supports_mid_conversation_system_messages":false},"vertex_ai/claude-haiku-5-5":{"mode":"chat","base_model":"claude-haiku-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai","provider":"vertex_ai","supports_native_effort":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_audio_input":false,"supports_audio_output":false,"supports_native_streaming":true,"supports_safeguards":true,"supports_reasoning_disable":true,"default_reasoning_effort":"medium","reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_output_config":true,"supports_mid_conversation_system_messages":false},"vertex_ai/claude-haiku-5-5@default":{"mode":"chat","base_model":"claude-haiku-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai","provider":"vertex_ai","supports_native_effort":true,"knowledge_cutoff":"2026-06-30","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_audio_input":false,"supports_audio_output":false,"supports_native_streaming":true,"supports_safeguards":true,"supports_reasoning_disable":true,"default_reasoning_effort":"medium","reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_output_config":true,"supports_mid_conversation_system_messages":false},"vertex_ai/claude-sonnet-5-5@default":{"mode":"chat","base_model":"claude-sonnet-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","supports_safeguards":true,"supports_web_search":true,"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false},"openai.gpt-6-luna":{"mode":"chat","base_model":"gpt-6-luna","max_input_tokens":1050000,"max_output_tokens":128000,"use_openai_responses_path":true,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supports_web_search":true,"provider":"bedrock_mantle","regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_parallel_function_calling":true,"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"web_search_options":false,"prompt_cache_key":false,"prompt_cache_retention":false,"verbosity":false,"store":false,"prompt_cache_options":false},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default"],"bedrock_mantle_base_path":"openai/v1","max_tokens":128000,"model_parameters":[{"id":"metadata","label":"Metadata","helpText":"A unique identifier representing your end-user.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"seed","label":"Seed","helpText":"(BETA) - System will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.","type":"number","range":{"min":0,"max":100},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"logprobs","label":"Logprobs","helpText":"Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"top_logprobs","label":"Top Logprobs","helpText":"An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability.","type":"number","range":{"min":0,"max":5,"step":1},"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"temperature","label":"Temperature","helpText":"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.","type":"number","range":{"min":0,"max":2,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"top_p","label":"Top P","helpText":"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.","type":"number","range":{"min":0,"max":1,"step":0.01},"default":1,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]},{"id":"max_completion_tokens","label":"Max Completion Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions"]},{"id":"max_output_tokens","label":"Max Output Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/responses"]},{"id":"max_tool_calls","label":"Max Tool Calls","helpText":"The maximum number of total calls to built-in tools that can be processed in a response.","type":"number","range":{"min":1,"max":128,"step":1},"supported_endpoints":["/v1/responses"]},{"id":"prompt_cache_key","label":"Prompt Cache Key","helpText":"Used by OpenAI to cache responses for similar requests to optimize your cache hit rates.","type":"text","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"prompt_cache_retention","label":"Prompt Cache Retention","helpText":"The retention policy for the prompt cache.","type":"select","options":[{"label":"In Memory","value":"in_memory"},{"label":"24 Hours","value":"24h"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"response_format","label":"Response Format","helpText":"The format of the response. If set to 'json', the response will be a JSON object with the keys 'choices' and 'logprobs'. If set to 'text', the response will be a string with the completion text. If set to 'Structured output', the response will match the supplied JSON schema.","type":"select","accesorKey":"type","default":{"type":"text"},"options":[{"label":"Text","value":"text"},{"label":"JSON","value":"json_object"},{"label":"Structured output","value":"json_schema"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"web_search","label":"Web Search","helpText":"Enable web search tool for real-time information retrieval.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"global.openai.gpt-6-sol":{"mode":"chat","base_model":"gpt-6-sol","provider":"bedrock","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"regional_endpoint_uplift_multiplier":1.1,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_image_input":true,"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_none_reasoning_effort":true,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_native_streaming":true,"supports_system_messages":true,"supports_web_search":true,"supports_file_search":true,"supports_computer_use":true,"supports_tool_search":true,"supports_service_tier":true,"source":"https://developers.openai.com/api/docs/models/gpt-6-sol","metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"supports_minimal_reasoning_effort":false,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"temperature":true,"top_p":true,"stop":true},"service_tiers":["auto","default"],"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":2048,"range":{"min":16,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]},{"id":"reasoning_effort","label":"Reasoning Effort","helpText":"This parameter controls how many reasoning tokens the model generates before producing a response","type":"select","options":[{"label":"None","value":"none"},{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"},{"label":"XHigh","value":"xhigh"},{"label":"Max","value":"max"}],"default":"medium","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"reasoning_summary","label":"Reasoning Summary","helpText":"Controls the level of detail in the reasoning summary output","type":"select","options":[{"label":"Auto","value":"auto"}],"supported_endpoints":["/v1/responses"]},{"id":"verbosity","label":"Verbosity","helpText":"Verbosity determines how many output tokens are generated. Lowering the number of tokens reduces overall latency.","type":"select","options":[{"label":"Low","value":"low"},{"label":"Medium","value":"medium"},{"label":"High","value":"high"}],"supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"promptTools","label":"Prompt Tools (Functions)","helpText":"A list of tools the model may call.","type":"select","supported_endpoints":["/v1/chat/completions","/v1/responses"]},{"id":"stream","label":"Stream","helpText":"The stream parameter in the API controls whether the response is sent in incremental updates, like tokenized data as it's generated, or as a complete result in one go.","type":"boolean","default":false,"supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/responses"]}]},"bedrock/claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_bedrock_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":1024,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrockFoundationModels/current/index.json","provider":"bedrock","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"supports_reasoning_disable":false,"supports_safeguards":true},"vertex_ai/claude-sonnet-5-5":{"mode":"chat","base_model":"claude-sonnet-5-5","regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":false,"supports_reasoning_disable":false,"supports_safeguards":true,"supports_web_search":true},"azure/gpt-realtime-2.1-2026-07-07":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"provider":"azure","base_model":"gpt-realtime-2.1"},"azure/gpt-realtime-2.1-mini-2026-07-07":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_reasoning":true,"provider":"azure","base_model":"gpt-realtime-2.1-mini"},"azure/gpt-realtime-translate-2026-05-06":{"max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/a-new-chapter-for-realtime-ai-reasoning-translation-and-real-time-transcription/4517124","supported_modalities":["audio"],"supported_output_modalities":["audio","text"],"supports_audio_input":true,"supports_audio_output":true,"supports_native_streaming":true,"provider":"azure","base_model":"gpt-realtime-translate"},"azure/gpt-realtime-translate-2026-05-07":{"max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://ai.azure.com/catalog/models/gpt-realtime-translate","supported_modalities":["audio"],"supported_output_modalities":["audio","text"],"supports_audio_input":true,"supports_audio_output":true,"supports_native_streaming":true,"provider":"azure","base_model":"gpt-realtime-translate"},"azure/gpt-realtime-whisper-2026-05-06":{"mode":"audio_transcription","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/gpt-realtime-whisper","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"max_input_tokens":32000,"max_output_tokens":4096,"provider":"azure","base_model":"gpt-realtime-whisper"},"azure/gpt-realtime-whisper-2026-05-07":{"mode":"audio_transcription","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/gpt-realtime-whisper","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"max_input_tokens":32000,"max_output_tokens":4096,"provider":"azure","base_model":"gpt-realtime-whisper"},"abliteration/abliterated-model":{"max_input_tokens":262144,"max_output_tokens":262134,"mode":"chat","source":"https://abliteration.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"reasoning_effort_levels":["none","minimal","low","medium","high","xhigh","max"],"provider":"abliteration","base_model":"abliterated-model"},"abliteration/abliterated-model-large":{"max_input_tokens":1000000,"max_output_tokens":999990,"mode":"chat","source":"https://abliteration.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["none","minimal","low","medium","high","xhigh","max"],"provider":"abliteration","base_model":"abliterated-model-large"},"abliteration/abliterated-model-large-v2":{"max_input_tokens":1000000,"max_output_tokens":999990,"mode":"chat","source":"https://abliteration.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_none_reasoning_effort":false,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["minimal","low","medium","high","xhigh","max"],"provider":"abliteration","base_model":"abliterated-model-large-v2"},"abliteration/abliterated-model-large-flash":{"max_input_tokens":1048576,"max_output_tokens":262144,"mode":"chat","source":"https://abliteration.ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"reasoning_effort_levels":["none","minimal","low","medium","high","xhigh","max"],"provider":"abliteration","base_model":"abliterated-model-large-flash"},"aiml/openai/gpt-image-2":{"metadata":{"notes":"OpenAI gpt-image-2 via AI/ML API - flagship multimodal image generation and editing model with reasoning and 2K output. output_cost_per_image is AI/ML's published medium-quality rate; like the other aiml image entries it is billed as a flat per-image price"},"mode":"image_generation","source":"https://docs.aimlapi.com/api-references/image-models/openai/gpt-image-2","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"aiml","base_model":"gpt-image-2"},"us.amazon.nova-canvas-v1:0":{"deprecation_date":"2026-09-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":2600,"mode":"image_generation","supports_nova_canvas_image_edit":true,"provider":"bedrock","base_model":"nova-canvas","is_deprecated":true},"amazon.nova-premier-v1:0":{"max_input_tokens":1000000,"max_output_tokens":10000,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"nova-premier"},"amazon.nova-2-sonic-v1:0":{"max_input_tokens":1000000,"max_output_tokens":64000,"mode":"realtime","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-sonic.md","supports_audio_input":true,"supports_audio_output":true,"provider":"bedrock","base_model":"nova-2-sonic"},"bedrock/claude-fable-latest":{"alias_of":"global.anthropic.claude-fable-5-2","base_model":"claude-fable"},"bedrock/claude-opus-latest":{"alias_of":"global.anthropic.claude-opus-5-5","base_model":"claude-opus"},"azure/claude-fable-latest":{"alias_of":"azure_ai/claude-fable-5-2","base_model":"claude-fable"},"azure/claude-mythos-5-1":{"deprecation_date":"2027-09-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":false,"supports_web_search":true,"prompt_cache_min_tokens":512,"source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","provider":"azure","base_model":"claude-mythos-5-1"},"azure/claude-opus-latest":{"alias_of":"azure_ai/claude-opus-5-5","base_model":"claude-opus"},"azure/gpt-astra-latest":{"alias_of":"azure/gpt-6-astra","base_model":"gpt-astra"},"azure/gpt-luna-latest":{"alias_of":"azure/gpt-6-luna","base_model":"gpt-luna"},"azure/gpt-sol-latest":{"alias_of":"azure/gpt-6.1-sol","base_model":"gpt-sol"},"azure/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/gpt-terra-latest":{"alias_of":"azure/gpt-6.1-terra","base_model":"gpt-terra"},"azure/gpt-image-2":{"deprecation_date":"2027-10-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"azure","base_model":"gpt-image-2","supports_prompt_caching":true},"azure/whisper":{"deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"audio_transcription","source":"https://management.azure.com/subscriptions/a88257a5-3323-4ef9-95bf-5681a46c9624/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","provider":"azure","base_model":"whisper"},"azure/gpt-mini-latest":{"alias_of":"azure/gpt-5.4-mini","base_model":"gpt-mini"},"azure/eu/gpt-4o":{"deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"gpt-4o"},"azure/eu/gpt-4o-mini":{"deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"gpt-4o-mini"},"azure/eu/o1":{"deprecation_date":"2026-11-19","successor_model":"azure/eu/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"o1"},"azure/eu/o3-mini":{"deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","base_model":"o3-mini"},"azure/gpt-audio":{"deprecation_date":"2027-03-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","provider":"azure","base_model":"gpt-audio"},"azure/gpt-audio-1.5":{"deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","supported_endpoints":["/v1/chat/completions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_output":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","base_model":"gpt-audio-1.5"},"azure/gpt-realtime":{"deprecation_date":"2027-03-02","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","provider":"azure","base_model":"gpt-realtime","supports_prompt_caching":true},"azure/gpt-realtime-1.5":{"deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-realtime-2.1-2026-07-07","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","provider":"azure","base_model":"gpt-realtime-1.5","supports_prompt_caching":true},"azure/gpt-realtime-1.5-2026-02-23":{"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-realtime-2.1-2026-07-07","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-1.5","supports_prompt_caching":true},"azure/gpt-realtime-2.1":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-2.1"},"azure/gpt-realtime-2.1-mini":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-2.1-mini"},"azure/gpt-realtime-mini":{"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","deprecation_date":"2027-06-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-mini","supports_prompt_caching":true},"azure/gpt-realtime-whisper":{"deprecation_date":"2027-05-06","max_input_tokens":32000,"max_output_tokens":4096,"mode":"audio_transcription","provider_specific_entry":{"realtime_ga_only":1},"source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/gpt-realtime-whisper","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"azure","base_model":"gpt-realtime-whisper"},"azure/gpt-5.3-chat":{"deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":111616,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-5.3-chat","is_deprecated":true},"azure/us/gpt-5.2-chat":{"deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-chat-latest","max_input_tokens":111616,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-5.2-chat","is_deprecated":true},"azure/us/gpt-5.3-chat":{"deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-chat-latest","max_input_tokens":111616,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-5.3-chat","is_deprecated":true},"azure/eu/gpt-5.2-chat":{"deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":111616,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-5.2-chat","is_deprecated":true},"azure/eu/gpt-5.3-chat":{"deprecation_date":"2026-06-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-chat-latest","max_input_tokens":111616,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-5.3-chat","is_deprecated":true},"azure/us/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/us/gpt-6.1-sol":{"deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-sol"},"azure/eu/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/eu/gpt-6.1-sol":{"deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-sol"},"azure/apac/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"apac/gpt-6.1-terra"},"azure/apac/gpt-6.1-sol":{"deprecation_date":"2028-03-11","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"apac/gpt-6.1-sol"},"azure/gpt-image-2.5-flare":{"deprecation_date":"2027-09-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"azure","base_model":"gpt-image-2.5-flare","supports_prompt_caching":true},"azure/gpt-image-2.5-sunburst":{"deprecation_date":"2027-09-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"azure","base_model":"gpt-image-2.5-sunburst","supports_prompt_caching":true},"azure/gpt-image-2-2026-04-21":{"deprecation_date":"2027-10-21","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"azure","base_model":"gpt-image-2","supports_prompt_caching":true},"azure/speech/azure-stt":{"audio_transcription_config":"azure_speech","mode":"audio_transcription","source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/speech-services/","supported_endpoints":["/v1/audio/transcriptions"],"provider":"azure","base_model":"speech/azure-stt"},"azure/us/gpt-4o":{"deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.1","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"gpt-4o"},"azure/us/gpt-4o-mini":{"deprecation_date":"2027-04-14","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"gpt-4o-mini"},"azure/us/o1":{"deprecation_date":"2026-11-19","successor_model":"azure/us/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"o1"},"azure/us/o3":{"deprecation_date":"2026-11-19","successor_model":"azure/us/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"o3"},"azure/us/o3-mini":{"deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"azure","base_model":"o3-mini"},"azure/us/o4-mini":{"deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/us/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"o4-mini"},"azure/FLUX.2-flex":{"deprecation_date":"2027-01-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":32000,"mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/black-forest-labs/","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"azure","base_model":"flux.2-flex"},"azure/kimi-latest":{"alias_of":"azure_ai/FW-Kimi-K3","base_model":"kimi"},"azure/minimax-latest":{"alias_of":"azure_ai/FW-MiniMax-M3","base_model":"minimax"},"azure/MAI-Image-2.5":{"mode":"image_generation","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"deprecation_date":"2026-10-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/MAI-Image-2.6","max_input_tokens":32000,"provider":"azure","base_model":"mai-image-2.5","is_deprecated":true},"azure/MAI-Image-2.5-Flash":{"mode":"image_generation","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"deprecation_date":"2026-10-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/MAI-Image-2.6-Flash","max_input_tokens":32000,"provider":"azure","base_model":"mai-image-2.5-flash","is_deprecated":true},"azure/MAI-Image-2.5-Pro":{"deprecation_date":"2026-10-01","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":32000,"mode":"image_generation","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-mai-image-2-5-pro-and-mai-voice-2-flash-in-microsoft-foundry/4539446","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","base_model":"mai-image-2.5-pro","is_deprecated":true},"azure/mistral-document-ai-2512":{"deprecation_date":"2027-02-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","mode":"ocr","supported_endpoints":["/v1/ocr"],"source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/mistral-ocr-4-0","provider":"azure","base_model":"mistral-document-ai"},"azure/mistral-ocr-4-0":{"mode":"ocr","supported_endpoints":["/v1/ocr"],"source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/mistral/","provider":"azure","base_model":"mistral-ocr-4-0"},"azure/Cohere-parse-v5":{"deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","mode":"ocr","source":"https://cohere.com/blog/parse","supported_endpoints":["/v1/ocr"],"provider":"azure","base_model":"parse-v5"},"azure/deepseek-pro-latest":{"alias_of":"azure_ai/deepseek-v4-pro","base_model":"deepseek-pro"},"azure/deepseek-flash-latest":{"alias_of":"azure_ai/deepseek-v4-flash","base_model":"deepseek-flash"},"bedrock/guardrails":{"mode":"guardrail","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"guardrails"},"black_forest_labs/flux-kontext-pro":{"mode":"image_edit","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-kontext-pro"},"black_forest_labs/flux-kontext-max":{"mode":"image_edit","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-kontext-max"},"black_forest_labs/flux-pro-1.0-fill":{"mode":"image_edit","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/edits"],"provider":"black_forest_labs","base_model":"flux-pro-1.0-fill"},"black_forest_labs/flux-pro-1.0-expand":{"mode":"image_edit","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/edits"],"provider":"black_forest_labs","base_model":"flux-pro-1.0-expand"},"black_forest_labs/flux-pro-1.1":{"mode":"image_generation","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-pro-1.1"},"black_forest_labs/flux-pro-1.1-ultra":{"mode":"image_generation","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-pro-1.1-ultra"},"black_forest_labs/flux-dev":{"mode":"image_generation","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-dev"},"black_forest_labs/flux-pro":{"mode":"image_generation","source":"https://bfl.ai/pricing","supported_endpoints":["/v1/images/generations"],"provider":"black_forest_labs","base_model":"flux-pro"},"cerebras/llama-3.1-70b":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"provider":"cerebras","base_model":"llama-3.1-70b"},"cerebras/llama-3.1-8b":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"provider":"cerebras","base_model":"llama-3.1-8b"},"cloudflare/clef-omni":{"max_input_tokens":64000,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/models/clef-omni/","supports_audio_input":true,"supports_video_input":true,"supports_vision":true,"provider":"cloudflare","base_model":"clef-omni","supports_decisions":true},"cloudflare/@cf/cloudflare/clef-omni":{"max_input_tokens":64000,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/models/clef-omni/","supports_audio_input":true,"supports_video_input":true,"supports_vision":true,"provider":"cloudflare","base_model":"clef-omni","supports_decisions":true},"cloudflare/clef":{"max_input_tokens":65536,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/","supports_vision":true,"provider":"cloudflare","base_model":"clef","supports_decisions":true},"cloudflare/clef-flash":{"max_input_tokens":65536,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/","supports_vision":true,"provider":"cloudflare","base_model":"clef-flash","supports_decisions":true},"cloudflare/@cf/cloudflare/clef":{"max_input_tokens":65536,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/","supports_vision":true,"provider":"cloudflare","base_model":"","supports_decisions":true},"cloudflare/@cf/cloudflare/clef-flash":{"max_input_tokens":65536,"mode":"decisions","source":"https://developers.cloudflare.com/workers-ai/platform/pricing/","supports_vision":true,"provider":"cloudflare","base_model":"clef-flash","supports_decisions":true},"cloudflare/clairic":{"max_input_tokens":65536,"mode":"extraction","source":"https://developers.cloudflare.com/workers-ai/models/clairic/","supports_vision":true,"provider":"cloudflare","base_model":"clairic"},"cloudflare/@cf/cloudflare/clairic":{"max_input_tokens":65536,"mode":"extraction","source":"https://developers.cloudflare.com/workers-ai/models/clairic/","supports_vision":true,"provider":"cloudflare","base_model":""},"eu.cohere.embed-v4:0":{"max_input_tokens":128000,"mode":"embedding","output_vector_size":1536,"supports_embedding_image_input":true,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"embed","supports_vision":true,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","sa-east-1"],"max_tokens":128000},"cohere/embed-v5.0-pro":{"max_input_tokens":128000,"mode":"embedding","output_vector_size":2048,"supports_embedding_image_input":true,"source":"https://cohere.com/pricing","provider":"cohere","base_model":"embed-v5.0-pro"},"cohere/embed-v5.0-fast":{"max_input_tokens":128000,"mode":"embedding","output_vector_size":2048,"supports_embedding_image_input":true,"source":"https://cohere.com/pricing","provider":"cohere","base_model":"embed-v5.0-fast"},"cohere/parse-v5.0":{"mode":"ocr","source":"https://cohere.com/blog/parse","supported_endpoints":["/v1/ocr"],"provider":"cohere","base_model":"parse"},"c4ai-aya-expanse-32b":{"max_input_tokens":128000,"max_output_tokens":4000,"mode":"chat","source":"https://docs.cohere.com/docs/models","provider":"cohere","base_model":"c4ai-aya-expanse-32b"},"c4ai-aya-vision-32b":{"max_input_tokens":16384,"max_output_tokens":4000,"mode":"chat","source":"https://docs.cohere.com/docs/aya-vision","supports_vision":true,"provider":"cohere","base_model":"c4ai-aya-vision-32b"},"command-a-reasoning-08-2025":{"max_input_tokens":288768,"max_output_tokens":32000,"mode":"chat","source":"https://docs.cohere.com/docs/command-a-reasoning","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":false,"provider":"cohere","base_model":"command-a"},"command-a-translate-08-2025":{"max_input_tokens":8992,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/command-a-translate","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","base_model":"command-a-translate"},"command-a-vision-07-2025":{"max_input_tokens":128000,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/command-a-vision","supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"cohere","base_model":"command-a-vision"},"command-r7b-arabic-02-2025":{"max_input_tokens":128000,"mode":"chat","source":"https://docs.cohere.com/docs/models","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"cohere","base_model":"command-r7b-arabic"},"north-mini-code-1.0":{"max_input_tokens":436000,"max_output_tokens":64000,"mode":"chat","source":"https://docs.cohere.com/docs/north-mini-code-1.0","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":false,"provider":"cohere","base_model":"north-mini-code-1.0"},"north-small-translate-1.0":{"max_input_tokens":32768,"max_output_tokens":16000,"mode":"chat","source":"https://docs.cohere.com/docs/north-small-translate-1.0","provider":"cohere","base_model":"north-small-translate-1.0"},"tiny-aya-earth":{"max_input_tokens":8192,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/tiny-aya","supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","base_model":"tiny-aya-earth"},"tiny-aya-fire":{"max_input_tokens":8192,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/tiny-aya","supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","base_model":"tiny-aya-fire"},"tiny-aya-global":{"max_input_tokens":8192,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/tiny-aya","supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","base_model":"tiny-aya-global"},"tiny-aya-water":{"max_input_tokens":8192,"max_output_tokens":8000,"mode":"chat","source":"https://docs.cohere.com/docs/tiny-aya","supports_function_calling":true,"supports_tool_choice":true,"provider":"cohere","base_model":"tiny-aya-water"},"coralbricks/glm-5.3-fast":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"coralbricks","base_model":"glm-5.3-fast"},"coralbricks/glm-5.3-fp4":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"coralbricks","base_model":"glm-5.3-fp4"},"coralbricks/glm-5.3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"coralbricks","base_model":"glm-5.3"},"coralbricks/deepseek-v4.1-flash-fast":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"coralbricks","base_model":"deepseek-v4.1-flash-fast"},"coralbricks/deepseek-v4.1-flash-fast-fp4":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"coralbricks","base_model":"deepseek-v4.1-flash-fast-fp4"},"coralbricks/deepseek-v4.1-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coralbricks.ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"coralbricks","base_model":"deepseek-v4.1-flash"},"computer-use-preview-2025-03-11":{"deprecation_date":"2026-07-23","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-terra","max_input_tokens":8192,"max_output_tokens":1024,"mode":"chat","fedramp_uplift_multiplier":1.1,"supported_endpoints":["/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://developers.openai.com/api/docs/models/computer-use-preview","provider":"openai","base_model":"computer-use","web_search_billing_unit":"per_request","is_deprecated":true},"dashscope/qwen4-max":{"max_input_tokens":991808,"max_output_tokens":131072,"mode":"chat","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"provider":"dashscope","base_model":"qwen4-max"},"dashscope/qwen-max-latest":{"alias_of":"dashscope/qwen3.8-max","base_model":"qwen-max"},"dashscope/qwen-flash-latest":{"alias_of":"dashscope/qwen3.8-flash","base_model":"qwen-flash"},"dashscope/qwen-image-2.0":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"dashscope","base_model":"qwen-image-2.0"},"dashscope/qwen-image-2.0-pro":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"dashscope","base_model":"qwen-image-2.0-pro"},"dashscope/qwen-image-3.0":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"dashscope","base_model":"qwen-image-3.0"},"dashscope/qwen-image-3.0-pro":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/models","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"dashscope","base_model":"qwen-image-3.0-pro"},"dashscope/qwen3-tts-flash":{"mode":"audio_speech","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/audio/speech"],"provider":"dashscope","base_model":"qwen3-tts-flash"},"dashscope/qwen-audio-3.1-tts-flash":{"mode":"audio_speech","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/audio/speech"],"provider":"dashscope","base_model":"qwen-audio-3.1-tts-flash"},"dashscope/qwen-audio-3.1-tts-plus":{"mode":"audio_speech","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/audio/speech"],"provider":"dashscope","base_model":"qwen-audio-3.1-tts-plus"},"dashscope/qwen-audio-3.0-tts-plus":{"mode":"audio_speech","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/audio/speech"],"provider":"dashscope","base_model":"qwen-audio-3.0-tts-plus"},"dashscope/qwen-audio-3.0-tts-flash":{"mode":"audio_speech","source":"https://www.alibabacloud.com/help/en/model-studio/model-pricing","supported_endpoints":["/v1/audio/speech"],"provider":"dashscope","base_model":"qwen-audio-3.0-tts-flash"},"dashscope/wan3.0-video":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 all-in-one video generation (text, first/last frame, multimodal reference, edit, extend). Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"wan3.0-video"},"dashscope/wan3.0-video-prime":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video-prime","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 fast tier. Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"wan3.0-video-prime"},"dashscope/happyhorse-1.1-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 text-to-video. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.1-t2v"},"dashscope/happyhorse-1.1-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 image-to-video from a first frame. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.1-i2v"},"dashscope/happyhorse-1.1-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.1 reference-to-video, 1 to 9 reference images. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.1-r2v"},"dashscope/happyhorse-1.0-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 text-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.0-t2v"},"dashscope/happyhorse-1.0-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 image-to-video from a first frame. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.0-i2v"},"dashscope/happyhorse-1.0-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.0 reference-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"happyhorse-1.0-r2v"},"dashscope/wan2.7-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-t2v","supported_modalities":["text","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 text-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"wan2.7-t2v"},"dashscope/wan2.7-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-i2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 image-to-video (first frame, first and last frame, continuation). No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"wan2.7-i2v"},"dashscope/wan2.7-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-r2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"Wan 2.7 reference-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"dashscope","base_model":"wan2.7-r2v"},"qwencloud/qwen-image-2.0":{"mode":"image_generation","source":"https://www.qwencloud.com/models","supported_endpoints":["/v1/images/generations"],"provider":"qwencloud","base_model":"qwen-image-2.0"},"qwencloud/qwen-image-2.0-pro":{"mode":"image_generation","source":"https://www.qwencloud.com/models","supported_endpoints":["/v1/images/generations"],"provider":"qwencloud","base_model":"qwen-image-2.0-pro"},"qwencloud/qwen-image-3.0":{"mode":"image_generation","source":"https://www.qwencloud.com/models","supported_endpoints":["/v1/images/generations"],"provider":"qwencloud","base_model":"qwen-image-3.0"},"qwencloud/qwen-image-3.0-pro":{"mode":"image_generation","source":"https://www.qwencloud.com/models","supported_endpoints":["/v1/images/generations"],"provider":"qwencloud","base_model":"qwen-image-3.0-pro"},"qwencloud/wan3.0-video":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 all-in-one video generation (text, first/last frame, multimodal reference, edit, extend). Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"wan3.0-video"},"qwencloud/wan3.0-video-prime":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video-prime","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 fast tier. Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"wan3.0-video-prime"},"qwencloud/happyhorse-1.1-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 text-to-video. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.1-t2v"},"qwencloud/happyhorse-1.1-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 image-to-video from a first frame. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.1-i2v"},"qwencloud/happyhorse-1.1-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.1 reference-to-video, 1 to 9 reference images. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.1-r2v"},"qwencloud/happyhorse-1.0-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 text-to-video. No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.0-t2v"},"qwencloud/happyhorse-1.0-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 image-to-video from a first frame. No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.0-i2v"},"qwencloud/happyhorse-1.0-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.0 reference-to-video. No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"happyhorse-1.0-r2v"},"qwencloud/wan2.7-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-t2v","supported_modalities":["text","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 text-to-video. No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"wan2.7-t2v"},"qwencloud/wan2.7-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-i2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 image-to-video (first frame, first and last frame, continuation). No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"wan2.7-i2v"},"qwencloud/wan2.7-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-r2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"Wan 2.7 reference-to-video. No 480P tier. Singapore list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwencloud","base_model":"wan2.7-r2v"},"qwen_ai_platform/qwen-image-2.0":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/models","supported_endpoints":["/v1/images/generations"],"provider":"qwen_ai_platform","base_model":"qwen-image-2.0"},"qwen_ai_platform/qwen-image-2.0-pro":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/models","supported_endpoints":["/v1/images/generations"],"provider":"qwen_ai_platform","base_model":"qwen-image-2.0-pro"},"qwen_ai_platform/qwen-image-3.0":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/models","supported_endpoints":["/v1/images/generations"],"provider":"qwen_ai_platform","base_model":"qwen-image-3.0"},"qwen_ai_platform/qwen-image-3.0-pro":{"mode":"image_generation","source":"https://www.alibabacloud.com/help/en/model-studio/models","supported_endpoints":["/v1/images/generations"],"provider":"qwen_ai_platform","base_model":"qwen-image-3.0-pro"},"qwen_ai_platform/wan3.0-video":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 all-in-one video generation (text, first/last frame, multimodal reference, edit, extend). Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"wan3.0-video"},"qwen_ai_platform/wan3.0-video-prime":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan3-0-video-prime","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 3.0 fast tier. Billed duration is usage.duration, which adds input video seconds for the edit and extend modes. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"wan3.0-video-prime"},"qwen_ai_platform/happyhorse-1.1-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 text-to-video. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.1-t2v"},"qwen_ai_platform/happyhorse-1.1-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.1 image-to-video from a first frame. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.1-i2v"},"qwen_ai_platform/happyhorse-1.1-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-1-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.1 reference-to-video, 1 to 9 reference images. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.1-r2v"},"qwen_ai_platform/happyhorse-1.0-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-t2v","supported_modalities":["text"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 text-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.0-t2v"},"qwen_ai_platform/happyhorse-1.0-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-i2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"HappyHorse 1.0 image-to-video from a first frame. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.0-i2v"},"qwen_ai_platform/happyhorse-1.0-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/happyhorse-1-0-r2v","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"HappyHorse 1.0 reference-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"happyhorse-1.0-r2v"},"qwen_ai_platform/wan2.7-t2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-t2v","supported_modalities":["text","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 text-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"wan2.7-t2v"},"qwen_ai_platform/wan2.7-i2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-i2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"first_frame"},"metadata":{"comment":"Wan 2.7 image-to-video (first frame, first and last frame, continuation). No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"wan2.7-i2v"},"qwen_ai_platform/wan2.7-r2v":{"mode":"video_generation","source":"https://www.alibabacloud.com/help/en/model-studio/wan2-7-r2v","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"dashscope_video_reference_input":"reference_image"},"metadata":{"comment":"Wan 2.7 reference-to-video. No 480P tier. China (Beijing) list price per output second, USD, matching this prefix's default host. The base rate is the 1080P default tier."},"provider":"qwen_ai_platform","base_model":"wan2.7-r2v"},"databricks/claude-fable-latest":{"alias_of":"databricks/databricks-claude-fable-5-2","base_model":"claude-fable"},"databricks/databricks-claude-opus-5-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_anthropic_thinking_payload":true,"supports_assistant_prefill":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_output_config":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"databricks","base_model":"claude-opus-5-5"},"databricks/claude-opus-latest":{"alias_of":"databricks/databricks-claude-opus-5-5","base_model":"claude-opus"},"databricks/system.ai.deepseek-v4-flash-0731":{"deprecation_date":"2026-11-05","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/retired-models-policy","mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/retired-models-policy","successor_model":"databricks/databricks-deepseek-v4-1-flash-1017","token_count_framing":97,"provider":"databricks","base_model":"system.ai.deepseek-v4-flash","is_deprecated":true},"databricks/glm-latest":{"alias_of":"databricks/databricks-glm-5-3","base_model":"glm"},"databricks/glm-flash-latest":{"alias_of":"databricks/databricks-glm-5-3-flash","base_model":"glm-flash"},"databricks/gpt-mini-latest":{"alias_of":"databricks/databricks-gpt-5-4-mini","base_model":"gpt-mini"},"databricks/kimi-latest":{"alias_of":"databricks/databricks-kimi-k3","base_model":"kimi"},"databricks/system.ai.kimi-k3":{"mode":"chat","token_count_framing":100,"provider":"databricks","base_model":"system.ai.kimi-k3"},"databricks/databricks-openjev-qwen35-4b":{"metadata":{"notes":"Pay-per-token Foundation Model API decision endpoint (OpenJev, Qwen 3.5 4B) queried through the System One API. The serving-endpoints API lists it at 0.6 DBUs per 1M input tokens (input_price on the foundation_model entry, read 2026-10-27); input_cost_per_token is that dbu cost * $0.070, the same conversion as the other databricks rows. Output is free. Databricks publishes no cache rates for it, so cache tokens bill at the input rate."},"mode":"decisions","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","provider":"databricks","base_model":"openjev-qwen35-4b","supports_decisions":true},"databricks/databricks-typesafeai-jev-1-14-0":{"metadata":{"notes":"TypeSafe AI Jev 1.14.0, provider-hosted by TypeSafe AI and served through Unity Gateway, queried through the System One API. The proprietary pricing page lists 'Jev 1.14' at 0.600 DBU per 1M input tokens; input_cost_per_token is that times $0.070 per DBU. Output is free. No cache rates are published, so cache tokens bill at the input rate."},"mode":"evaluation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supports_vision":true,"provider":"databricks","base_model":"typesafeai-jev-1-14-0"},"databricks/ai_decide":{"metadata":{"notes":"The ai_decide AI Function (Beta), called through POST /api/2.0/ai-functions/ai-decide. Its response carries no token usage, so LiteLLM records no spend for it."},"mode":"decisions","source":"https://docs.databricks.com/api/ai-functions/v1/ai-decide","provider":"databricks","base_model":"ai-decide","supports_decisions":true},"deepgram/aura":{"metadata":{"calculation":"$0.0150/1000 characters = $0.000015 per character","note":"Aura-1 text-to-speech voices (aura-<voice>-<lang>), pay as you go"},"source":"https://deepgram.com/pricing","provider":"deepgram","base_model":"aura"},"deepgram/aura-2":{"metadata":{"calculation":"$0.030/1000 characters = $0.00003 per character","note":"Aura-2 text-to-speech voices (aura-2-<voice>-<lang>), pay as you go"},"source":"https://deepgram.com/pricing","provider":"deepgram","base_model":"aura-2"},"deepgram/streaming/nova-3":{"metadata":{"calculation":"$0.0077/60 seconds = $0.00012833 per second","note":"Nova-3 monolingual streaming, pay as you go","original_pricing_per_minute":0.0077},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/nova-3"},"deepgram/streaming/nova-3-medical":{"metadata":{"calculation":"$0.0077/60 seconds = $0.00012833 per second","note":"Nova-3 Medical streaming, launch price from Deepgram's 2025-03-03 announcement. deepgram.com/pricing does not list Nova-3 Medical, and this matches its current Nova-3 monolingual streaming pay as you go rate","original_pricing_per_minute":0.0077},"mode":"audio_transcription","source":"https://www.businesswire.com/news/home/20250303470439/en","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/nova-3-medical"},"deepgram/streaming/nova-3-multilingual":{"metadata":{"calculation":"$0.0092/60 seconds = $0.00015333 per second","note":"Nova-3 multilingual (language=multi) streaming, pay as you go","original_pricing_per_minute":0.0092},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/nova-3-multilingual"},"deepgram/streaming/redact":{"metadata":{"calculation":"$0.0020/60 seconds = $0.00003333 per second","note":"Redaction add-on (redact query param), streaming, pay as you go","original_pricing_per_minute":0.002},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/redact"},"deepgram/streaming/keyterm":{"metadata":{"calculation":"$0.0013/60 seconds = $0.00002167 per second","note":"Keyterm Prompting add-on (keyterm query param), streaming, pay as you go","original_pricing_per_minute":0.0013},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/keyterm"},"deepgram/streaming/detect_entities":{"metadata":{"calculation":"$0.0017/60 seconds = $0.00002833 per second","note":"Entity Detection add-on (detect_entities query param), streaming, pay as you go","original_pricing_per_minute":0.0017},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/detect-entities"},"deepgram/streaming/diarize":{"metadata":{"calculation":"$0.0020/60 seconds = $0.00003333 per second","note":"Speaker Diarization add-on (diarize / diarize_model query params), streaming, pay as you go","original_pricing_per_minute":0.002},"mode":"audio_transcription","source":"https://deepgram.com/pricing","supported_endpoints":["/v1/listen"],"provider":"deepgram","base_model":"streaming/diarize"},"webiq/search":{"mode":"search","provider":"webiq","base_model":"search"},"sail/moonshotai/Kimi-K3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"kimi-k3"},"sail/zai-org/GLM-5.3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"glm-5.3"},"sail/zai-org/GLM-5.3-Flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"glm-5.3-flash"},"sail/deepseek-ai/DeepSeek-V4-Pro-0813":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"deepseek-v4-pro"},"sail/deepseek-ai/DeepSeek-V4-Flash-0731":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"deepseek-v4-flash"},"sail/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"deepseek-v4.1-flash"},"sail/moonshotai/Kimi-K2.6":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"kimi-k2.6"},"sail/google/gemma-4-31B-it":{"max_input_tokens":256000,"max_output_tokens":256000,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"gemma-4-31b-it"},"sail/nvidia/Gemma-4-31B-IT-NVFP4":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"gemma-4-31b-it-nvfp4"},"sail/google/gemma-4-12B-it":{"max_input_tokens":16384,"max_output_tokens":16384,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"gemma-4-12b-it"},"sail/openai/gpt-oss-120b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"gpt-oss-120b"},"sail/Qwen/Qwen3.6-35B-A3B":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://docs.sailresearch.com/models","provider":"sail","base_model":"qwen3.6-35b-a3b"},"serper/search":{"mode":"search","metadata":{"notes":"Serper Google Search API. Pricing: $1.00/1k queries (Starter), $0.75/1k (Standard), $0.50/1k (Scale), $0.30/1k (Ultimate)."},"provider":"serper","base_model":"search"},"apiserpent/search":{"mode":"search","metadata":{"notes":"APISerpent quick search (/api/search/quick), multi-engine (Google, Bing, Yahoo, DuckDuckGo). Pricing: $0.60/1k searches."},"provider":"apiserpent","base_model":"search"},"apiserpent/deep_search":{"mode":"search","metadata":{"notes":"APISerpent deep search (/api/search), multi-engine (Google, Bing, Yahoo, DuckDuckGo). Pricing: $0.60/1k searches."},"provider":"apiserpent","base_model":"deep-search"},"agentcore/search":{"mode":"search","metadata":{"notes":"Web Search on Amazon Bedrock AgentCore, billed by AWS on the gateway"},"provider":"agentcore","base_model":"search"},"bing_grounding/search":{"mode":"search","metadata":{"notes":"Grounding with Bing Search (G1 SKU): $14 per 1,000 transactions. Tokens for the Foundry model deployment that runs the grounded search are billed separately on that deployment."},"provider":"bing_grounding","base_model":"search"},"deepseek/search":{"mode":"search","provider":"deepseek","base_model":"search","supports_native_effort":true},"tinyfish/search":{"mode":"search","metadata":{"notes":"TinyFish Search API"},"provider":"tinyfish","base_model":"search"},"nimble/search":{"mode":"search","metadata":{"notes":"Nimble Search API pay-as-you-go list price: $5 per 1,000 searches, up to 100 results per search. Volume plans price differently."},"provider":"nimble","base_model":"search"},"elevenlabs/eleven_v4":{"metadata":{"calculation":"$0.08/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven v4 - most emotive TTS model with audio tags and 90+ languages"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-v4"},"elevenlabs/eleven_v4_turbo":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven v4 Turbo - low-latency v4 tuned for real-time and agents"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-v4-turbo"},"elevenlabs/eleven_v3_conversational":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven v3 Conversational - low-latency v3 tuned for realtime conversation"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-v3-conversational"},"elevenlabs/eleven_flash_v2_5":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven Flash v2.5 - low-latency multilingual TTS model"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-flash-v2-5"},"elevenlabs/eleven_flash_v2":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven Flash v2 - low-latency English-only TTS model"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-flash-v2"},"elevenlabs/eleven_turbo_v2_5":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven Turbo v2.5 - multilingual TTS model"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-turbo-v2-5"},"elevenlabs/eleven_turbo_v2":{"metadata":{"calculation":"$0.04/1000 characters (API usage pricing, billed in USD per character)","notes":"ElevenLabs Eleven Turbo v2 - English-only TTS model"},"mode":"audio_speech","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/speech"],"provider":"elevenlabs","base_model":"eleven-turbo-v2"},"fal_ai/bytedance/seedance-2.5/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/text-to-video"},"fal_ai/bytedance/seedance-latest/text-to-video":{"alias_of":"fal_ai/bytedance/seedance-2.5/text-to-video","base_model":"seedance-latest/text-to-video"},"fal_ai/bytedance/seedance-2.5/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/image-to-video"},"fal_ai/bytedance/seedance-latest/image-to-video":{"alias_of":"fal_ai/bytedance/seedance-2.5/image-to-video","base_model":"seedance-latest/image-to-video"},"fal_ai/bytedance/seedance-2.5/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/reference-to-video"},"fal_ai/bytedance/seedance-latest/reference-to-video":{"alias_of":"fal_ai/bytedance/seedance-2.5/reference-to-video","base_model":"seedance-latest/reference-to-video"},"fal_ai/minimax/h3/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 2k rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/text-to-video"},"fal_ai/minimax/h3/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/reference-to-video"},"fal_ai/bytedance/seedance-2.0/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/text-to-video","metadata":{"comment":"fal bills $0.014 per 1k tokens (480p/720p/1080p) and $0.008 per 1k tokens (4k) with tokens = h*w*seconds*24/1024; 480p and 4k rates derived from that formula at 854x480 and 3840x2160"},"supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/text-to-video"},"fal_ai/bytedance/seedance-2.0/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/image-to-video","metadata":{"comment":"fal bills $0.014 per 1k tokens (480p/720p/1080p) and $0.008 per 1k tokens (4k) with tokens = h*w*seconds*24/1024; 480p and 4k rates derived from that formula at 854x480 and 3840x2160"},"supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/image-to-video"},"fal_ai/bytedance/seedance-2.0/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/reference-to-video","metadata":{"comment":"fal bills $0.014 per 1k tokens (480p/720p/1080p) and $0.008 per 1k tokens (4k) with tokens = h*w*seconds*24/1024; 480p and 4k rates derived from that formula at 854x480 and 3840x2160"},"supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/reference-to-video"},"fal_ai/fal-ai/nano-banana":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"nano-banana"},"fal_ai/fal-ai/gemini-25-flash-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-25-flash-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"gemini-25-flash-image"},"fal_ai/fal-ai/nano-banana-2":{"metadata":{"comment":"priced by the request's resolution field (0.5K, 1K default, 2K, 4K); the web search and high thinking surcharges are not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana-2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"nano-banana-2"},"fal_ai/fal-ai/nano-banana-pro":{"metadata":{"comment":"priced by the request's resolution field (1K default, 2K, 4K); the web search surcharge is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana-pro","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"nano-banana-pro"},"fal_ai/google/nano-banana-2.1":{"metadata":{"comment":"fal bills this endpoint by tokens and reports the charge in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are fal's representative costs by the request's resolution field (1K default, 2K, 4K) for a short prompt, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/google/nano-banana-2.1","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"nano-banana-2.1"},"fal_ai/google/nano-banana-2.1/edit":{"metadata":{"comment":"fal bills this endpoint by tokens and reports the charge in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are fal's representative costs by the request's resolution field (1K default, 2K, 4K) for a short prompt and two 1K input images, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/google/nano-banana-2.1/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nano-banana-2.1/edit"},"fal_ai/google/nano-banana-latest":{"alias_of":"fal_ai/google/nano-banana-2.1","base_model":"nano-banana"},"fal_ai/google/nano-banana-latest/edit":{"alias_of":"fal_ai/google/nano-banana-2.1/edit","base_model":"nano-banana-latest/edit"},"fal_ai/openai/gpt-image-2":{"metadata":{"notes":""},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2"},"fal_ai/gpt-image-2":{"metadata":{"notes":""},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2"},"fal_ai/openai/gpt-image-2/edit":{"metadata":{"notes":"Editing endpoint of gpt-image-2 on fal.ai, reached through the image generation path with fal's image_urls param since /v1/images/edits is not wired for fal_ai. Prices include one input image and live in keyed entries fal_ai/{quality}/{width}-x-{height}/openai/gpt-image-2/edit. This flat entry is the fallback for the default edit request (quality=high, image_size=auto, inferred from the input image, priced as 1024x768 high)"},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2/edit","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"edit"},"fal_ai/openai/gpt-image-2.5/flare/text-to-image":{"metadata":{"notes":"OpenAI gpt-image-2.5 (flare) served through fal.ai. fal publishes deterministic per-image prices per size and quality, mirrored as keyed entries fal_ai/{quality}/{width}-x-{height}/openai/gpt-image-2.5/flare/text-to-image that the fal_ai cost calculator picks from the request params. This flat entry is the fallback for the default request (quality=high, image_size=landscape_4_3 at 1024x768). quality=auto is priced as high"},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2.5/flare/text-to-image","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2.5/flare/text-to-image"},"fal_ai/openai/gpt-image-2.5/flare/edit":{"metadata":{"notes":"Editing endpoint of gpt-image-2.5 (flare) on fal.ai, reachable through /v1/images/edits or the image generation path with fal's image_urls param. Prices include one input image and live in keyed entries fal_ai/{quality}/{width}-x-{height}/openai/gpt-image-2.5/flare/edit. This flat entry is the fallback for the default edit request (quality=high, image_size=auto, inferred from the input image, priced as 1024x768 high)"},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2.5/flare/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2.5/flare/edit"},"fal_ai/openai/gpt-image-2.5/sunburst/text-to-image":{"metadata":{"notes":"OpenAI gpt-image-2.5 (sunburst) served through fal.ai. fal publishes deterministic per-image prices per size and quality, mirrored as keyed entries fal_ai/{quality}/{width}-x-{height}/openai/gpt-image-2.5/sunburst/text-to-image that the fal_ai cost calculator picks from the request params. This flat entry is the fallback for the default request (quality=high, image_size=landscape_4_3 at 1024x768). quality=auto is priced as high"},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2.5/sunburst/text-to-image","supported_endpoints":["/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2.5/sunburst/text-to-image"},"fal_ai/openai/gpt-image-2.5/sunburst/edit":{"metadata":{"notes":"Editing endpoint of gpt-image-2.5 (sunburst) on fal.ai, reachable through /v1/images/edits or the image generation path with fal's image_urls param. Prices include one input image and live in keyed entries fal_ai/{quality}/{width}-x-{height}/openai/gpt-image-2.5/sunburst/edit. This flat entry is the fallback for the default edit request (quality=high, image_size=auto, inferred from the input image, priced as 1024x768 high)"},"mode":"image_generation","source":"https://fal.ai/models/openai/gpt-image-2.5/sunburst/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gpt-image-2.5/sunburst/edit"},"fal_ai/fal-ai/flux/dev":{"metadata":{"notes":"fal bills fal-ai/flux/dev at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/dev","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux/dev"},"fal_ai/fal-ai/trellis":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/trellis","metadata":{"comment":"image-to-3D, returns a GLB mesh; served through the /fal_ai pass-through route"},"provider":"fal_ai","base_model":"trellis"},"fal_ai/fal-ai/trellis-2":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/trellis-2","metadata":{"comment":"image-to-3D, returns a GLB mesh; priced by the request's resolution field (default 1024); served through the /fal_ai pass-through route"},"provider":"fal_ai","base_model":"trellis-2"},"fal_ai/fal-ai/flux-lora-depth":{"metadata":{"notes":"fal bills fal-ai/flux-lora-depth at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora-depth","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-lora-depth"},"fireworks_ai/accounts/fireworks/routers/minimax-latest":{"max_input_tokens":512000,"max_output_tokens":512000,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/minimax"},"fireworks_ai/minimax-latest":{"alias_of":"fireworks_ai/accounts/fireworks/models/minimax-m3","base_model":"minimax"},"gemini-3-pro-image":{"max_input_tokens":65536,"max_output_tokens":32768,"mode":"image_generation","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-3-pro-image"},"gemini-3.1-flash-image":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"image_generation","search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-3.1-flash-image"},"gemini-3.1-flash-image-preview":{"max_input_tokens":65536,"max_output_tokens":32768,"mode":"image_generation","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-3.1-flash-image"},"gemini-live-2.5-flash-native-audio":{"deprecation_date":"2026-12-13","deprecation_source":"https://cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions","successor_model":"gemini-3.8-live","max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/vertex_ai/live","/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":false,"supports_vision":true,"supports_web_search":true,"gemini_native_audio":true,"provider":"vertex_ai","base_model":"gemini-live-2.5-flash-native-audio"},"vertex_ai/gemini-3.8-flash-cyber":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","regional_endpoint_uplift_multiplier":1.1,"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":false,"supports_url_context":false,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"provider":"vertex_ai","base_model":"gemini-3.8-flash-cyber"},"vertex_ai/gemini-3.8-live":{"gemini_audio_only_live":true,"max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/vertex_ai/live","/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_response_schema":false,"supports_vision":true,"supports_web_search":true,"provider":"vertex_ai","base_model":"gemini-3.8-live"},"gemini-embedding-2-preview":{"max_input_tokens":8192,"mode":"embedding","output_vector_size":3072,"source":"https://cloud.google.com/vertex-ai/generative-ai/pricing","supports_embedding_image_input":true,"uses_embed_content":true,"provider":"vertex_ai","base_model":"gemini-embedding-2"},"gemini/gemini-3.1-flash-image-preview":{"deprecation_date":"2026-06-25","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-nano-banana-2.1","max_input_tokens":65536,"max_output_tokens":65536,"mode":"image_generation","rpm":1000,"tpm":4000000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits"],"supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"web_search_billing_unit":"per_query","provider":"gemini","base_model":"gemini-3.1-flash-image","is_deprecated":true},"gemini/gemini-omni-flash-preview":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"video_generation","rpm":2000,"source":"https://ai.google.dev/gemini-api/docs/deprecations","supported_endpoints":["/v1beta/interactions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","video"],"supports_audio_input":true,"supports_reasoning":true,"supports_system_messages":false,"supports_video_input":true,"supports_vision":true,"tpm":800000,"deprecation_date":"2026-10-22","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-omni-1.2-flash","provider":"gemini","base_model":"gemini-omni-flash-preview","supports_function_calling":false,"supports_parallel_function_calling":false,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_tool_choice":false,"is_deprecated":true},"gemini-3.8-flash-cyber":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_output":false,"supports_audio_input":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_minimal_reasoning_effort":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_system_messages":true,"supports_tool_choice":false,"supports_url_context":false,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_native_streaming":true,"provider":"vertex_ai","base_model":"gemini-3.8-flash-cyber"},"gemini/veo-3.1-lite-generate-preview":{"deprecation_date":"2026-10-22","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-omni-1.2-flash","max_input_tokens":1024,"mode":"video_generation","source":"https://ai.google.dev/gemini-api/docs/deprecations","supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"gemini","base_model":"veo-3.1-lite-generate","is_deprecated":true},"github_copilot/claude-fable-5":{"deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/claude-fable-5.2","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":true,"provider":"github_copilot","base_model":"claude-fable-5"},"github_copilot/claude-fable-5.1":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":true,"provider":"github_copilot","base_model":"claude-fable-5-1"},"github_copilot/claude-fable-latest":{"alias_of":"github_copilot/claude-fable-5.2","base_model":"claude-fable"},"github_copilot/claude-haiku-5.5":{"max_input_tokens":128000,"max_output_tokens":16000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":false,"provider":"github_copilot","base_model":"claude-haiku-5-5"},"github_copilot/claude-haiku-latest":{"alias_of":"github_copilot/claude-haiku-5.5","base_model":"claude-haiku"},"github_copilot/claude-opus-4.8":{"deprecation_date":"2026-11-16","deprecation_source":"https://github.blog/changelog/2026-11-16-github-copilot-models-deprecated-on-november-16-2026","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/claude-opus-5.5","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"claude-opus-4-8","is_deprecated":true},"github_copilot/claude-opus-4.8-fast":{"deprecation_date":"2026-11-16","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"claude-opus-4-8-fast","is_deprecated":true},"github_copilot/claude-opus-5":{"deprecation_date":"2026-11-16","deprecation_source":"https://github.blog/changelog/2026-11-16-github-copilot-models-deprecated-on-november-16-2026","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/claude-opus-5.5","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"claude-opus-5","is_deprecated":true},"github_copilot/claude-opus-5.5":{"max_input_tokens":200000,"max_output_tokens":128000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":true,"provider":"github_copilot","base_model":"claude-opus-5-5"},"github_copilot/claude-opus-latest":{"alias_of":"github_copilot/claude-opus-5.5","base_model":"claude-opus"},"github_copilot/claude-sonnet-5":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"claude-sonnet-5"},"github_copilot/claude-sonnet-5.5":{"max_input_tokens":200000,"max_output_tokens":128000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":true,"provider":"github_copilot","base_model":"claude-sonnet-5-5"},"github_copilot/claude-sonnet-latest":{"alias_of":"github_copilot/claude-sonnet-5.5","base_model":"claude-sonnet"},"github_copilot/gemini-3.7-flash":{"deprecation_date":"2026-10-19","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_vision":true,"provider":"github_copilot","base_model":"gemini-3.7-flash","is_deprecated":true},"github_copilot/gemini-3.8-flash":{"deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":200000,"max_output_tokens":65536,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/gemini-4-hydrogen","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_vision":true,"provider":"github_copilot","base_model":"gemini-3.8-flash"},"github_copilot/gpt-5.4":{"deprecation_date":"2026-10-19","deprecation_source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","max_input_tokens":272000,"max_output_tokens":128000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.4","is_deprecated":true},"github_copilot/gpt-5.4-mini":{"deprecation_date":"2026-10-19","max_input_tokens":272000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.4-mini","is_deprecated":true},"github_copilot/gpt-5.5":{"deprecation_date":"2026-10-19","deprecation_source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/model-deprecation-history.yml","max_input_tokens":272000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.5","is_deprecated":true},"github_copilot/gpt-5.6-luna":{"max_input_tokens":200000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.6-luna"},"github_copilot/gpt-5.6-sol":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.6-sol"},"github_copilot/gpt-5.6-terra":{"deprecation_date":"2026-11-16","deprecation_source":"https://github.blog/changelog/2026-11-16-github-copilot-models-deprecated-on-november-16-2026","max_input_tokens":272000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.6-terra","is_deprecated":true},"github_copilot/gpt-6-astra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6-astra"},"github_copilot/gpt-astra-latest":{"alias_of":"github_copilot/gpt-6-astra","base_model":"gpt-astra"},"github_copilot/gpt-6-luna":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6-luna"},"github_copilot/gpt-luna-latest":{"alias_of":"github_copilot/gpt-6-luna","base_model":"gpt-luna"},"github_copilot/gpt-6-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6-sol"},"github_copilot/gpt-6.1-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6.1-sol"},"github_copilot/gpt-sol-latest":{"alias_of":"github_copilot/gpt-6.1-sol","base_model":"gpt-sol"},"github_copilot/kimi-k3":{"deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":917504,"max_output_tokens":131072,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"kimi-k3"},"github_copilot/kimi-latest":{"alias_of":"github_copilot/kimi-k3","base_model":"kimi"},"github_copilot/mai-code-1.1-flash":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"mai-code-1.1-flash"},"chatgpt/gpt-6-sol":{"mode":"responses","supported_endpoints":["/v1/chat/completions","/v1/responses"],"max_input_tokens":922000,"max_output_tokens":128000,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6-sol"},"chatgpt/gpt-6-luna":{"mode":"responses","supported_endpoints":["/v1/chat/completions","/v1/responses"],"max_input_tokens":922000,"max_output_tokens":128000,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6-luna"},"chatgpt/gpt-6-astra":{"mode":"responses","supported_endpoints":["/v1/chat/completions","/v1/responses"],"max_input_tokens":922000,"max_output_tokens":128000,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6-astra"},"chatgpt/gpt-6.1-sol":{"mode":"responses","supported_endpoints":["/v1/chat/completions","/v1/responses"],"max_input_tokens":922000,"max_output_tokens":128000,"source":"https://developers.openai.com/api/docs/pricing","supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6.1-sol"},"gpt-image-2":{"deprecation_date":"2027-02-10","deprecation_source":"https://developers.openai.com/api/docs/deprecations","mode":"image_generation","source":"https://developers.openai.com/api/docs/pricing","successor_model":"gpt-image-2.6-sunburst","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","base_model":"gpt-image-2","web_search_billing_unit":"per_request"},"gpt-image-2-2026-04-21":{"deprecation_date":"2027-02-10","deprecation_source":"https://developers.openai.com/api/docs/deprecations","mode":"image_generation","source":"https://developers.openai.com/api/docs/pricing","successor_model":"gpt-image-2.6-sunburst","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_vision":true,"supports_pdf_input":true,"provider":"openai","base_model":"gpt-image-2","web_search_billing_unit":"per_request"},"gpt-5.3-chat-latest":{"deprecation_date":"2026-08-10","deprecation_source":"https://developers.openai.com/api/docs/deprecations","successor_model":"gpt-5.6-sol","max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-5.3-chat-latest","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"openai","base_model":"gpt-5.3-chat","web_search_billing_unit":"per_request","is_deprecated":true},"gpt-realtime-1.5":{"max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","base_model":"gpt-realtime-1.5","web_search_billing_unit":"per_request"},"gpt-realtime-2":{"max_input_tokens":128000,"max_output_tokens":32000,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","base_model":"gpt-realtime-2","web_search_billing_unit":"per_request"},"gpt-realtime-2.1":{"max_input_tokens":128000,"max_output_tokens":32000,"mode":"realtime","regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","base_model":"gpt-realtime-2.1","web_search_billing_unit":"per_request"},"gpt-realtime-2.1-mini":{"max_input_tokens":128000,"max_output_tokens":32000,"mode":"realtime","regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"openai","base_model":"gpt-realtime-2.1-mini","web_search_billing_unit":"per_request"},"groq/canopylabs/orpheus-v1-english":{"max_input_tokens":4000,"max_output_tokens":50000,"mode":"audio_speech","source":"https://console.groq.com/docs/model/canopylabs/orpheus-v1-english","provider":"groq","base_model":"canopylabs/orpheus-v1-english"},"groq/canopylabs/orpheus-arabic-saudi":{"max_input_tokens":4000,"max_output_tokens":50000,"mode":"audio_speech","source":"https://console.groq.com/docs/models","provider":"groq","base_model":"canopylabs/orpheus-arabic-saudi"},"hostinger_router/gpt-5.2":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"source":"https://router.hostinger.com/v1/models","provider":"hostinger_router","base_model":"gpt-5.2"},"hostinger_router/claude-sonnet-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"supports_response_schema":true,"supported_endpoints":["/v1/chat/completions","/v1/messages"],"source":"https://router.hostinger.com/v1/models","provider":"hostinger_router","base_model":"claude-sonnet-5"},"hostinger_router/text-embedding-3-large":{"max_input_tokens":8191,"mode":"embedding","source":"https://router.hostinger.com/v1/models","provider":"hostinger_router","base_model":"text-embedding-3-large"},"hostinger_router/text-embedding-3-small":{"max_input_tokens":8191,"mode":"embedding","source":"https://router.hostinger.com/v1/models","provider":"hostinger_router","base_model":"text-embedding-3-small"},"jina-reranker-v3.5":{"max_input_tokens":131072,"mode":"rerank","source":"https://api.jina.ai/v1/models","provider":"jina_ai","base_model":"jina-reranker"},"jina-reranker-m0":{"max_input_tokens":10240,"mode":"rerank","source":"https://api.jina.ai/v1/models","provider":"jina_ai","base_model":"jina-reranker-m0"},"inferenceapis/deepseek-ai/DeepSeek-V3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v3"},"inferenceapis/deepseek-ai/DeepSeek-V4-Flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-flash"},"inferenceapis/deepseek-ai/DeepSeek-V4-Flash-0731":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-flash"},"inferenceapis/deepseek-ai/DeepSeek-V4-Pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-pro"},"inferenceapis/deepseek-ai/DeepSeek-V4-Pro-0813":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-pro"},"inferenceapis/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4.1-flash"},"inferenceapis/deepseek-chat":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-chat"},"inferenceapis/deepseek-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-flash"},"inferenceapis/deepseek-reasoner":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-r1"},"inferenceapis/deepseek-v3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v3"},"inferenceapis/deepseek-v4-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-flash"},"inferenceapis/deepseek-v4-pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4-pro"},"inferenceapis/deepseek-v4.1-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"deepseek-v4.1-flash"},"inferenceapis/glm-5.3":{"max_input_tokens":1048575,"max_output_tokens":1048575,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"glm-5.3"},"inferenceapis/glm-5.3-flash":{"max_input_tokens":1048575,"max_output_tokens":1048575,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"glm-5.3-flash"},"inferenceapis/gpt-oss-120b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"gpt-oss-120b"},"inferenceapis/gpt-oss-20b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"gpt-oss-20b"},"inferenceapis/kimi-k2.7-code":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"kimi-k2.7-code"},"inferenceapis/llama-3.1-8b-instant":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.1-8b-instant"},"inferenceapis/llama-3.3-70b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.3-70b"},"inferenceapis/llama-3.3-70b-instruct":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.3-70b-instruct"},"inferenceapis/llama-3.3-70b-versatile":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.3-70b-versatile"},"inferenceapis/meta-llama/Llama-3.3-70B-Instruct":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.3-70b-instruct"},"inferenceapis/meta-llama/Llama-3.3-70B-Instruct-Turbo":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"llama-3.3-70b-instruct-turbo"},"inferenceapis/minimax-m3":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"minimax-m3"},"inferenceapis/MiniMaxAI/MiniMax-M3":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"minimax-m3"},"inferenceapis/moonshotai/Kimi-K2.7-Code":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"kimi-k2.7-code"},"inferenceapis/openai/gpt-oss-120b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"gpt-oss-120b"},"inferenceapis/openai/gpt-oss-20b":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"gpt-oss-20b"},"inferenceapis/Qwen/Qwen3-VL-235B-A22B-Instruct":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":true,"provider":"inferenceapis","base_model":"qwen3-vl-235b-a22b-instruct"},"inferenceapis/Qwen/Qwen3.8-Flash":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"qwen3.8-flash"},"inferenceapis/qwen3-vl":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":true,"provider":"inferenceapis","base_model":"qwen3-vl"},"inferenceapis/qwen3-vl-235b":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":false,"supports_tool_choice":true,"supports_vision":true,"provider":"inferenceapis","base_model":"qwen3-vl-235b"},"inferenceapis/qwen3.8-flash":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"qwen3.8-flash"},"inferenceapis/zai-org/GLM-5.3":{"max_input_tokens":1048575,"max_output_tokens":1048575,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"glm-5.3"},"inferenceapis/zai-org/GLM-5.3-Flash":{"max_input_tokens":1048575,"max_output_tokens":1048575,"mode":"chat","source":"https://inferenceapis.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"inferenceapis","base_model":"glm-5.3-flash"},"meta/muse-ember-1.0":{"max_input_tokens":1048576,"max_output_tokens":524288,"mode":"chat","source":"https://dev.meta.ai/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","base_model":"muse-ember-1.0"},"meta/muse-voice-transcribe-1.0":{"mode":"audio_transcription","source":"https://dev.meta.ai/docs/speech-to-text","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"meta","base_model":"muse-voice-transcribe-1.0"},"meta/muse-wonder-1.0":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://dev.meta.ai/docs/pricing-rate-limits","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"meta","base_model":"muse-wonder-1.0"},"minimax/MiniMax-M3.1-Flash":{"mode":"chat","source":"https://platform.minimax.io/docs/guides/pricing-paygo","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_none_reasoning_effort":false,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_vision":true,"max_input_tokens":1000000,"provider":"minimax","base_model":"minimax-m3.1-flash"},"minimax/minimax-latest":{"alias_of":"minimax/MiniMax-M3","base_model":"minimax"},"minimax/MiniMax-H3":{"mode":"video_generation","source":"https://platform.minimax.io/docs/pricing/overview#models/video","provider_specific_entry":{"minimax_free_input_images":5},"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v2/video_generation"],"metadata":{"comment":"V2 async task API. Pay-as-you-go list price $0.08/s at 768P and $0.13/s at 2K, plus $0.04 per input image beyond the first 5. Audio input is free. Reference video input is rejected, since MiniMax bills its duration only after the create call."},"provider":"minimax","base_model":"MiniMax-H3","supports_image_input":true,"supports_video_input":true,"supports_audio_input":true},"minimax/MiniMax-H3-Max":{"mode":"video_generation","source":"https://platform.minimax.io/docs/pricing/overview#models/video","provider_specific_entry":{"minimax_free_input_images":2},"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v2/video_generation"],"metadata":{"comment":"V2 async task API, fast tier. Pay-as-you-go list price $0.05/s at 480P and $0.08/s at 768P; no 2K. Plus $0.074 per input image beyond the first 2. Audio input is free. Reference video input is rejected, since MiniMax bills its duration only after the create call."},"provider":"minimax","base_model":"MiniMax-H3-Max","supports_image_input":true,"supports_video_input":true,"supports_audio_input":true},"bedrock/ap-southeast-2/mistral.voxtral-mini-3b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/ap-southeast-2/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-mini-3b"},"bedrock/ap-southeast-2/mistral.voxtral-small-24b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/ap-southeast-2/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-small-24b"},"bedrock/ap-northeast-1/mistral.voxtral-mini-3b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/ap-northeast-1/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-mini-3b"},"bedrock/ap-northeast-1/mistral.voxtral-small-24b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/ap-northeast-1/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-small-24b"},"bedrock/eu-west-2/mistral.voxtral-mini-3b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/eu-west-2/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-mini-3b"},"bedrock/eu-west-2/mistral.voxtral-small-24b-2507":{"deprecation_date":"2027-03-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/eu-west-2/index.json","supports_audio_input":true,"supports_system_messages":true,"supports_native_structured_output":true,"provider":"bedrock","base_model":"voxtral-small-24b"},"mistral/voxtral-mini-transcribe-realtime-latest":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-realtime-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-realtime"},"mistral/voxtral-mini-tts-latest":{"mode":"audio_speech","source":"https://docs.mistral.ai/models/model-cards/voxtral-tts-26-03","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":true,"provider":"mistral","base_model":"voxtral-mini-tts"},"mistral/voxtral-mini-transcribe-3":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-mini-transcribe-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-3"},"mistral/voxtral-mini-transcribe-3-0":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-mini-transcribe-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-3-0"},"mistral/voxtral-small-transcribe-3":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-small-transcribe-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-small-transcribe-3"},"mistral/voxtral-small-transcribe-3-0":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-small-transcribe-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-small-transcribe-3-0"},"mistral/voxtral-mini-transcribe-realtime-3":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-mini-transcribe-realtime-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-realtime-3"},"mistral/voxtral-mini-transcribe-realtime-3-0":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/voxtral-mini-transcribe-realtime-3-0","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-realtime-3-0"},"mistral/voxtral-mini-tts-3":{"mode":"audio_speech","source":"https://docs.mistral.ai/models/voxtral-mini-tts-3-0","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":true,"provider":"mistral","base_model":"voxtral-mini-tts-3"},"mistral/voxtral-mini-tts-3-0":{"mode":"audio_speech","source":"https://docs.mistral.ai/models/voxtral-mini-tts-3-0","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":true,"provider":"mistral","base_model":"voxtral-mini-tts-3-0"},"mistral/deepseek-v4-1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://docs.mistral.ai/models/deepseek-v4-flash-1017","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","base_model":"deepseek-v4-1-flash","reasoning_effort_levels":["none","low","high","xhigh","max"]},"mistral/zai-glm-5-3-flash":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://docs.mistral.ai/models/zai-glm-5-3-flash","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","base_model":"glm-5-3-flash"},"mistral/mistral-ocr-4-0":{"deprecation_date":"2026-09-30","mode":"ocr","supported_endpoints":["/v1/ocr","/v1/batch"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","base_model":"mistral-ocr-4-0","is_deprecated":true},"mistral/mistral-ocr-4-1":{"mode":"ocr","source":"https://docs.mistral.ai/models/model-cards/ocr-4-1","supported_endpoints":["/v1/ocr","/v1/batch"],"provider":"mistral","base_model":"mistral-ocr-4-1"},"mistral/mistral-ocr-2512":{"mode":"ocr","supported_endpoints":["/v1/ocr","/v1/batch"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","base_model":"mistral-ocr"},"mistral/mistral-large-4":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","reasoning_effort_levels":["low","medium","high"],"source":"https://docs.mistral.ai/models/mistral-large-4-0","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","base_model":"mistral-large-4"},"mistral/mistral-large-4-0":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","reasoning_effort_levels":["low","medium","high"],"source":"https://docs.mistral.ai/models/mistral-large-4-0","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"mistral","base_model":"mistral-large-4-0"},"moonshot/kimi-k3.1-pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://platform.kimi.ai/docs/pricing/chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"moonshot","base_model":"kimi-k3.1-pro"},"moonshot/kimi-k3.1-pro-highspeed":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://platform.kimi.ai/docs/pricing/chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"moonshot","base_model":"kimi-k3.1-pro-highspeed"},"morph/morph-kimik3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-kimik3"},"morph/morph-kimik3-fast":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-kimik3-fast"},"morph/morph-glm53-744b":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","provider":"morph","base_model":"morph-glm53-744b"},"morph/morph-glm53flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-glm53flash"},"morph/morph-dsv41flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-dsv41flash"},"morph/morph-dsv41flash-1017":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-dsv41flash"},"morph/morph-dsv41flash-abliterated":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","supports_vision":true,"provider":"morph","base_model":"morph-dsv41flash-abliterated"},"morph/morph-dsv4flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.morphllm.com/pricing","provider":"morph","base_model":"morph-dsv4flash"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://tokenfactory.nebius.com/api/public/models_info","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"nebius","base_model":"deepseek-v4.1-flash"},"nebius/moonshotai/Kimi-K3.1-Pro":{"max_input_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"kimi-k3.1-pro"},"nebius/Qwen/Qwen3.8-27B":{"max_input_tokens":262144,"mode":"chat","source":"https://tokenfactory.nebius.com/api/public/models_info","supports_function_calling":true,"supports_reasoning":true,"provider":"nebius","base_model":"qwen3.8-27b"},"oci/xai.grok-4.20-0309-non-reasoning":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","reasoning_effort_levels":[],"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"provider":"oci","base_model":"grok-4.20-0309-non"},"oci/xai.grok-4.20-0309-reasoning":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","reasoning_effort_levels":[],"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_reasoning":true,"provider":"oci","base_model":"grok-4.20"},"oci/xai.grok-4.20-non-reasoning":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","reasoning_effort_levels":[],"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"provider":"oci","base_model":"grok-4.20-non"},"oci/xai.grok-4.20-reasoning":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","reasoning_effort_levels":[],"source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_reasoning":true,"provider":"oci","base_model":"grok-4.20"},"oci/openai.gpt-oss-120b":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"provider":"oci","base_model":"gpt-oss-120b"},"oci/openai.gpt-oss-20b":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","source":"https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing","supports_function_calling":true,"supports_native_streaming":true,"supports_reasoning":true,"provider":"oci","base_model":"gpt-oss-20b"},"parallel_ai/decision":{"max_input_tokens":32000,"mode":"evaluation","source":"https://docs.parallel.ai/getting-started/pricing#decisions-api","supported_endpoints":["/v1/decisions","/v1/systemone"],"provider":"parallel_ai","base_model":"decision"},"parallel_ai/search-fast":{"mode":"search","provider":"parallel_ai","base_model":"search-fast"},"parallel_ai/search-turbo":{"mode":"search","provider":"parallel_ai","base_model":"search-turbo"},"perplexity/pplx-decider-v1-27b":{"max_input_tokens":262143,"mode":"decisions","source":"https://docs.perplexity.ai/docs/decisions/quickstart.md","provider":"perplexity","base_model":"pplx-decider-v1-27b","supports_decisions":true,"max_tokens":262143,"supported_modalities":["text","image"],"supported_output_modalities":["text"]},"perplexity/pplx-decider-v1.1-27b":{"max_input_tokens":262143,"mode":"decisions","source":"https://docs.perplexity.ai/docs/decisions/quickstart.md","provider":"perplexity","base_model":"pplx-decider-v1.1-27b","supports_decisions":true,"max_tokens":262143,"supported_modalities":["text","image"],"supported_output_modalities":["text"]},"perplexity/pplx-decider-v1.2":{"max_input_tokens":262144,"mode":"evaluation","source":"https://docs.perplexity.ai/docs/decisions/quickstart","provider":"perplexity","base_model":"pplx-decider"},"perplexity/pplx-decider-latest":{"max_input_tokens":262144,"mode":"evaluation","source":"https://docs.perplexity.ai/docs/decisions/quickstart","provider":"perplexity","base_model":"pplx-decider"},"perplexity/pplx-decider-flash-v1.2":{"max_input_tokens":262144,"mode":"evaluation","source":"https://docs.perplexity.ai/docs/decisions/quickstart","provider":"perplexity","base_model":"pplx-decider-flash"},"perplexity/pplx-decider-flash-latest":{"max_input_tokens":262144,"mode":"evaluation","source":"https://docs.perplexity.ai/docs/decisions/quickstart","provider":"perplexity","base_model":"pplx-decider-flash"},"perplexity/perplexity/deepseek-v4-pro-0813":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"deepseek-v4-pro"},"reducto/parse-legacy":{"mode":"ocr","source":"https://reducto.ai/pricing","supported_endpoints":["/v1/ocr"],"provider":"reducto","base_model":"parse-legacy"},"reducto/parse-v3":{"mode":"ocr","source":"https://reducto.ai/pricing","supported_endpoints":["/v1/ocr"],"provider":"reducto","base_model":"parse-v3"},"rerank-v4.0-fast":{"max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://cohere.com/pricing","provider":"cohere","base_model":"rerank-v4.0-fast"},"rerank-v4.0-pro":{"max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://cohere.com/pricing","provider":"cohere","base_model":"rerank-v4.0-pro"},"rerank-v5.0-fast":{"max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://cohere.com/blog/rerank-5","provider":"cohere","base_model":"rerank-v5.0-fast"},"rerank-v5.0-pro":{"max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://cohere.com/blog/rerank-5","provider":"cohere","base_model":"rerank-v5.0-pro"},"nvidia_nim/01-ai/yi-large":{"mode":"chat","source":"https://build.nvidia.com/01-ai/yi-large.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"01-ai/yi-large"},"nvidia_nim/adept/fuyu-8b":{"source":"https://build.nvidia.com/adept/fuyu-8b.md","provider":"nvidia_nim","base_model":"adept/fuyu-8b"},"nvidia_nim/ai21labs/jamba-1.5-large-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"ai21labs/jamba-1.5-large-instruct"},"nvidia_nim/aisingapore/sea-lion-7b-instruct":{"mode":"chat","source":"https://build.nvidia.com/aisingapore/sea-lion-7b-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"sea-lion-7b-instruct"},"nvidia_nim/bigcode/starcoder2-15b":{"mode":"completion","source":"https://build.nvidia.com/bigcode/starcoder2-15b.md","supported_endpoints":["/v1/completions"],"provider":"nvidia_nim","base_model":"bigcode/starcoder2-15b"},"nvidia_nim/databricks/dbrx-instruct":{"mode":"chat","source":"https://build.nvidia.com/databricks/dbrx-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"dbrx-instruct"},"nvidia_nim/deepseek-ai/deepseek-coder-6.7b-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"deepseek-coder-6.7b-instruct"},"nvidia_nim/deepseek-ai/deepseek-v4.1-flash":{"max_output_tokens":1048576,"mode":"chat","source":"https://build.nvidia.com/deepseek-ai/deepseek-v4.1-flash.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"deepseek-v4.1-flash"},"nvidia_nim/google/codegemma-1.1-7b":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"codegemma-1.1-7b"},"nvidia_nim/google/codegemma-7b":{"mode":"chat","source":"https://build.nvidia.com/google/codegemma-7b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"codegemma-7b"},"nvidia_nim/google/deplot":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"deplot"},"nvidia_nim/google/diffusiongemma-26b-a4b-it":{"mode":"chat","source":"https://build.nvidia.com/google/diffusiongemma-26b-a4b-it.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"diffusiongemma-26b-a4b-it"},"nvidia_nim/google/gemma-2b":{"mode":"chat","source":"https://build.nvidia.com/google/gemma-2b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"gemma-2b"},"nvidia_nim/google/gemma-3-12b-it":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"gemma-3-12b-it"},"nvidia_nim/google/gemma-3-4b-it":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"gemma-3-4b-it"},"nvidia_nim/google/gemma-4-31b-it":{"mode":"chat","source":"https://build.nvidia.com/google/gemma-4-31b-it.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"gemma-4-31b-it"},"nvidia_nim/google/recurrentgemma-2b":{"mode":"chat","source":"https://build.nvidia.com/google/recurrentgemma-2b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"recurrentgemma-2b"},"nvidia_nim/ibm/granite-3.0-3b-a800m-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"granite-3.0-3b-a800m-instruct"},"nvidia_nim/ibm/granite-3.0-8b-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"granite-3.0-8b-instruct"},"nvidia_nim/ibm/granite-34b-code-instruct":{"mode":"chat","source":"https://build.nvidia.com/ibm/granite-34b-code-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"granite-34b-code-instruct"},"nvidia_nim/ibm/granite-8b-code-instruct":{"mode":"chat","source":"https://build.nvidia.com/ibm/granite-8b-code-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"granite-8b-code-instruct"},"nvidia_nim/meta/codellama-70b":{"mode":"chat","source":"https://build.nvidia.com/meta/codellama-70b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"codellama-70b"},"nvidia_nim/meta/llama-3.2-90b-vision-instruct":{"mode":"chat","source":"https://build.nvidia.com/meta/llama-3.2-90b-vision-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"llama-3.2-90b-vision-instruct"},"nvidia_nim/meta/llama-guard-4-12b":{"mode":"chat","source":"https://build.nvidia.com/meta/llama-guard-4-12b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"llama-guard-4-12b"},"nvidia_nim/meta/llama2-70b":{"mode":"chat","source":"https://build.nvidia.com/meta/llama2-70b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"llama-2-70b"},"nvidia_nim/meta/muse-glimmer-30b":{"max_output_tokens":16384,"mode":"chat","source":"https://build.nvidia.com/meta/muse-glimmer-30b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"muse-glimmer-30b"},"nvidia_nim/microsoft/kosmos-2":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"kosmos-2"},"nvidia_nim/microsoft/phi-3-vision-128k-instruct":{"source":"https://build.nvidia.com/microsoft/phi-3-vision-128k-instruct.md","provider":"nvidia_nim","base_model":"phi-3-vision-128k-instruct"},"nvidia_nim/microsoft/phi-3.5-moe-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"phi-3.5-moe-instruct"},"nvidia_nim/minimaxai/minimax-m3.1-flash":{"mode":"chat","source":"https://build.nvidia.com/minimaxai/minimax-m3.1-flash.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"provider":"nvidia_nim","base_model":"minimax-m3.1-flash"},"nvidia_nim/mistralai/codestral-22b-instruct-v0.1":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"codestral-22b-instruct"},"nvidia_nim/mistralai/mistral-7b-instruct-v0.3":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"mistral-7b-instruct"},"nvidia_nim/mistralai/mistral-large":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"mistral-large"},"nvidia_nim/mistralai/mistral-large-2-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"mistral-large-2-instruct"},"nvidia_nim/mistralai/mistral-large-4-1t-a52b":{"max_output_tokens":262144,"mode":"chat","source":"https://build.nvidia.com/mistralai/mistral-large-4-1t-a52b.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"nvidia_nim","base_model":"mistral-large-4-1t-a52b"},"nvidia_nim/mistralai/mixtral-8x22b-v0.1":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"mixtral-8x22b"},"nvidia_nim/moonshotai/kimi-k2.6":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"kimi-k2.6"},"nvidia_nim/nv-mistralai/mistral-nemo-12b-instruct":{"mode":"chat","source":"https://build.nvidia.com/nv-mistralai/mistral-nemo-12b-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nv-mistralai/mistral-nemo-12b-instruct"},"nvidia_nim/nvidia/ai-synthetic-video-detector":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"ai-synthetic-video-detector"},"nvidia_nim/nvidia/cosmos-reason2-8b":{"max_output_tokens":4096,"mode":"chat","source":"https://build.nvidia.com/nvidia/cosmos-reason2-8b/modelcard","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"cosmos-reason2-8b"},"nvidia_nim/nvidia/embed-qa-4":{"mode":"embedding","source":"https://build.nvidia.com/nvidia/embed-qa-4.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"embed-qa-4"},"nvidia_nim/nvidia/ising-calibration-1.5-31b":{"mode":"chat","source":"https://build.nvidia.com/nvidia/ising-calibration-1.5-31b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"ising-calibration-1.5-31b"},"nvidia_nim/nvidia/llama-3.1-nemoguard-8b-content-safety":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemoguard-8b-content-safety"},"nvidia_nim/nvidia/llama-3.1-nemoguard-8b-topic-control":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemoguard-8b-topic-control"},"nvidia_nim/nvidia/llama-3.1-nemotron-51b-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemotron-51b-instruct"},"nvidia_nim/nvidia/llama-3.1-nemotron-70b-instruct":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemotron-70b-instruct"},"nvidia_nim/nvidia/llama-3.1-nemotron-safety-guard-8b-v3":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemotron-safety-guard-8b-v3"},"nvidia_nim/nvidia/llama-3.1-nemotron-ultra-253b-v1":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.1-nemotron-ultra-253b-v1"},"nvidia_nim/nvidia/llama-3.2-nemoretriever-1b-vlm-embed-v1":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.2-nemoretriever-1b-vlm-embed-v1"},"nvidia_nim/nvidia/llama-3.2-nv-embedqa-1b-v1":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3.2-nv-embedqa-1b-v1"},"nvidia_nim/nvidia/llama-nemotron-embed-vl-1b-v2":{"mode":"embedding","source":"https://build.nvidia.com/nvidia/llama-nemotron-embed-vl-1b-v2.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"llama-nemotron-embed-vl-1b-v2"},"nvidia_nim/nvidia/llama3-chatqa-1.5-70b":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"llama-3-chatqa-1.5-70b"},"nvidia_nim/nvidia/mistral-nemo-minitron-8b-8k-instruct":{"mode":"chat","source":"https://build.nvidia.com/nvidia/mistral-nemo-minitron-8b-8k-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"mistral-nemo-minitron-8b-8k-instruct"},"nvidia_nim/nvidia/nemotron-3-embed-1b":{"mode":"embedding","source":"https://build.nvidia.com/nvidia/nemotron-3-embed-1b.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"nemotron-3-embed-1b"},"nvidia_nim/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-3-nano-omni-30b-a3b"},"nvidia_nim/nvidia/nemotron-3-super-120b-a12b":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-3-super-120b-a12b"},"nvidia_nim/nvidia/nemotron-3-ultra-550b-a55b":{"max_output_tokens":32768,"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-3-ultra-550b-a55b"},"nvidia_nim/nvidia/nemotron-3.5-bolt-e4b":{"max_output_tokens":65536,"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3.5-bolt-e4b.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"nvidia_nim","base_model":"nemotron-3.5-bolt-e4b"},"nvidia_nim/nvidia/nemotron-3.5-content-safety":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3.5-content-safety.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-3.5-content-safety"},"nvidia_nim/nvidia/nemotron-3.5-embed-1b":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"nemotron-3.5-embed-1b"},"nvidia_nim/nvidia/nemotron-3.5-embed-8b":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"nemotron-3.5-embed-8b"},"nvidia_nim/nvidia/nemotron-3.5-lightning-30b-a3b":{"max_output_tokens":32768,"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3.5-lightning-30b-a3b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-3.5-lightning-30b-a3b"},"nvidia_nim/nvidia/nemotron-3.5-super-vl-120b-a12b":{"max_output_tokens":131072,"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-3.5-super-vl-120b-a12b.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"nvidia_nim","base_model":"nemotron-3.5-super-vl-120b-a12b"},"nvidia_nim/nvidia/nemotron-4-340b-instruct":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-4-340b-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-4-340b-instruct"},"nvidia_nim/nvidia/nemotron-4-340b-reward":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-4-340b-reward.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-4-340b-reward"},"nvidia_nim/nvidia/nemotron-nano-3-30b-a3b":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"nemotron-nano-3-30b-a3b"},"nvidia_nim/nvidia/nemotron-parse":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-parse.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-parse"},"nvidia_nim/nvidia/nemotron-parse-2.0":{"mode":"chat","source":"https://build.nvidia.com/nvidia/nemotron-parse-2.0.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"nemotron-parse-2.0"},"nvidia_nim/nvidia/neva-22b":{"source":"https://build.nvidia.com/nvidia/neva-22b.md","provider":"nvidia_nim","base_model":"neva-22b"},"nvidia_nim/nvidia/nv-embedqa-mistral-7b-v2":{"mode":"embedding","source":"https://build.nvidia.com/nvidia/nv-embedqa-mistral-7b-v2.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"nv-embedqa-mistral-7b-v2"},"nvidia_nim/nvidia/nvclip":{"mode":"embedding","source":"https://build.nvidia.com/nvidia/nvclip.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"nvclip"},"nvidia_nim/nvidia/riva-translate-4b-instruct":{"mode":"chat","source":"https://build.nvidia.com/nvidia/riva-translate-4b-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"riva-translate-4b-instruct"},"nvidia_nim/nvidia/riva-translate-4b-instruct-v2":{"max_output_tokens":4096,"mode":"chat","source":"https://build.nvidia.com/nvidia/riva-translate-4b-instruct-v2.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"riva-translate-4b-instruct-v2"},"nvidia_nim/nvidia/vila":{"source":"https://build.nvidia.com/nvidia/vila.md","provider":"nvidia_nim","base_model":"vila"},"nvidia_nim/openai/gpt-oss-20b":{"max_output_tokens":4096,"mode":"chat","source":"https://build.nvidia.com/openai/gpt-oss-20b.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"provider":"nvidia_nim","base_model":"gpt-oss-20b"},"nvidia_nim/poolside/laguna-xs-2.1":{"max_output_tokens":16384,"mode":"chat","source":"https://build.nvidia.com/poolside/laguna-xs-2.1.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"poolside/laguna-xs-2.1"},"nvidia_nim/snowflake/arctic-embed-l":{"mode":"embedding","source":"https://build.nvidia.com/snowflake/arctic-embed-l.md","supported_endpoints":["/v1/embeddings"],"provider":"nvidia_nim","base_model":"arctic-embed-l"},"nvidia_nim/stepfun-ai/step-5-preview":{"max_output_tokens":1048576,"mode":"chat","source":"https://build.nvidia.com/stepfun-ai/step-5-preview.md","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"nvidia_nim","base_model":"stepfun-ai/step-5"},"nvidia_nim/writer/palmyra-creative-122b":{"mode":"chat","source":"https://build.nvidia.com/writer/palmyra-creative-122b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"writer/palmyra-creative-122b"},"nvidia_nim/writer/palmyra-fin-70b-32k":{"mode":"chat","source":"https://build.nvidia.com/writer/palmyra-fin-70b-32k.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"writer/palmyra-fin-70b-32k"},"nvidia_nim/writer/palmyra-med-70b":{"mode":"chat","source":"https://build.nvidia.com/writer/palmyra-med-70b.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"writer/palmyra-med-70b"},"nvidia_nim/writer/palmyra-med-70b-32k":{"mode":"chat","source":"https://build.nvidia.com/writer/palmyra-med-70b-32k.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"writer/palmyra-med-70b-32k"},"nvidia_nim/z-ai/glm-5.3-flash":{"source":"https://integrate.api.nvidia.com/v1/models","provider":"nvidia_nim","base_model":"glm-5.3-flash"},"nvidia_nim/zyphra/zamba2-7b-instruct":{"mode":"chat","source":"https://build.nvidia.com/zyphra/zamba2-7b-instruct.md","supported_endpoints":["/v1/chat/completions"],"provider":"nvidia_nim","base_model":"zyphra/zamba2-7b-instruct"},"scx-ai/DeepSeek-V4.1-flash":{"max_input_tokens":1000000,"max_output_tokens":375000,"mode":"chat","source":"https://scx.ai/models/deepseek-v4-1-flash","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"scx-ai","base_model":"deepseek-v4.1-flash"},"scx-ai/GLM-5.3":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://scx.ai/models/glm-5-3","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"provider":"scx-ai","base_model":"glm-5.3"},"scx-ai/Kimi-K3":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","source":"https://scx.ai/models/kimi-k3","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":false,"provider":"scx-ai","base_model":"kimi-k3"},"us.stability.stable-conservative-upscale-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-conservative-upscale"},"us.stability.stable-creative-upscale-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-creative-upscale"},"us.stability.stable-fast-upscale-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-fast-upscale"},"us.stability.stable-outpaint-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-outpaint"},"us.stability.stable-image-control-sketch-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-control-sketch"},"us.stability.stable-image-control-structure-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-control-structure"},"us.stability.stable-image-erase-object-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-erase-object"},"us.stability.stable-image-inpaint-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-inpaint"},"us.stability.stable-image-remove-background-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-remove-background"},"us.stability.stable-image-search-recolor-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-search-recolor"},"us.stability.stable-image-search-replace-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-search-replace"},"us.stability.stable-image-style-guide-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-image-style-guide"},"us.stability.stable-style-transfer-v1:0":{"max_input_tokens":77,"mode":"image_edit","source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"stable-style-transfer"},"you_com/search":{"mode":"search","provider":"you_com","base_model":"search"},"together_ai/MiniMaxAI/MiniMax-M3.1-Flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://www.together.ai/pricing","supports_prompt_caching":true,"provider":"together_ai","base_model":"minimax-m3.1-flash"},"together_ai/deepseek-pro-latest":{"alias_of":"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813","base_model":"deepseek-pro"},"together_ai/meta-models/Muse-Spark-Open-816B-A42B":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://docs.together.ai/docs/serverless-models","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"together_ai","base_model":"models/muse-spark-open-816b-a42b"},"together_ai/kimi-latest":{"alias_of":"together_ai/moonshotai/Kimi-K3","base_model":"kimi"},"together_ai/glm-latest":{"alias_of":"together_ai/zai-org/GLM-5.3","base_model":"glm"},"together_ai/glm-flash-latest":{"alias_of":"together_ai/zai-org/GLM-5.3-Flash","base_model":"glm-flash"},"transcribe/StartTranscriptionJob":{"mode":"audio_transcription","source":"https://aws.amazon.com/transcribe/pricing/","metadata":{"notes":"Amazon Transcribe standard batch transcription, billed per second of audio with no minimum. Same rate in every region of the AWS Price List offer file for transcribe (checked 2026-09-17)"},"provider":"transcribe","base_model":"starttranscriptionjob"},"vispark/vision-small":{"max_input_tokens":1000000,"max_output_tokens":65536,"mode":"chat","source":"https://api.lab.vispark.in/v1/models","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vispark","base_model":"vision-small"},"vispark/vision-medium":{"max_input_tokens":1000000,"max_output_tokens":65536,"mode":"chat","source":"https://api.lab.vispark.in/v1/models","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vispark","base_model":"vision-medium"},"vispark/vision-large":{"max_input_tokens":1000000,"max_output_tokens":65536,"mode":"chat","source":"https://api.lab.vispark.in/v1/models","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vispark","base_model":"vision-large"},"vercel_ai_gateway/openai/gpt-6-luna-decisions":{"max_input_tokens":1050000,"mode":"evaluation","source":"https://ai-gateway.vercel.sh/v1/models","provider":"vercel_ai_gateway","base_model":"gpt-6-luna-decisions"},"vercel_ai_gateway/typesafe-ai/jev":{"max_input_tokens":32000,"metadata":{"notes":"Vercel AI Gateway model list entry typesafe-ai/jev, read 2026-11-09: input 0.000000042 per token, output 0, context window 32000. A direct /v1/decisions probe with a 768x768 input_image answered 200 on 2026-11-10, while a 96x96 one answered 400 invalid request body."},"mode":"evaluation","source":"https://ai-gateway.vercel.sh/v1/models","supported_modalities":["text","image"],"supports_box_questions":true,"supports_vision":true,"provider":"vercel_ai_gateway","base_model":"typesafe-ai/jev"},"vercel_ai_gateway/typesafe-ai/jev-1.13.0":{"max_input_tokens":32000,"metadata":{"notes":"Vercel AI Gateway model list entry typesafe-ai/jev-1.13.0, read 2026-11-09: input 0.000000042 per token, output 0, context window 32000. Direct /v1/decisions probes on 2026-11-10 asking whether a 768x768 picture is a smooth gradient answered 0.99 for a gradient and 0.12 for solid gray."},"mode":"evaluation","source":"https://ai-gateway.vercel.sh/v1/models","supported_modalities":["text","image"],"supports_vision":true,"provider":"vercel_ai_gateway","base_model":"typesafe-ai/jev-1.13.0"},"vercel_ai_gateway/typesafe-ai/jev-1.14.0":{"max_input_tokens":32000,"metadata":{"notes":"Vercel AI Gateway model list entry typesafe-ai/jev-1.14.0, read 2026-11-09: input 0.000000042 per token, output 0, context window 32000. Direct /v1/decisions probes on 2026-11-10 asking whether a 768x768 picture is a smooth gradient answered 0.84 for a gradient and 0.13 for solid gray."},"mode":"evaluation","source":"https://ai-gateway.vercel.sh/v1/models","supported_modalities":["text","image"],"supports_box_questions":true,"supports_vision":true,"provider":"vercel_ai_gateway","base_model":"typesafe-ai/jev-1.14.0"},"vertex_ai/chirp_3":{"metadata":{"calculation":"$0.016/60 seconds = $0.00026667 per second","original_pricing_per_minute":0.016},"mode":"audio_transcription","source":"https://cloud.google.com/speech-to-text/pricing","supported_endpoints":["/v1/audio/transcriptions","/v1/realtime"],"provider":"vertex_ai","base_model":"chirp-3"},"vertex_ai/chirp_2":{"metadata":{"calculation":"$0.016/60 seconds = $0.00026667 per second","original_pricing_per_minute":0.016},"mode":"audio_transcription","source":"https://cloud.google.com/speech-to-text/pricing","supported_endpoints":["/v1/audio/transcriptions","/v1/realtime"],"provider":"vertex_ai","base_model":"chirp-2"},"vertex_ai/claude-mythos-5-2":{"regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-11-09","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"claude-mythos-5-2","max_tokens":128000,"provider_specific_entry":{"us":1.1},"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_output_config":true,"supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true,"supports_speed":true,"supports_fast_mode":true},"vertex_ai/claude-fable-latest":{"alias_of":"vertex_ai/claude-fable-5-2","base_model":"claude-fable"},"vertex_ai/claude-mythos-5-2@default":{"regional_endpoint_uplift_multiplier":1.1,"supports_mid_conversation_system":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":512,"deprecation_date":"2027-11-09","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"claude-mythos-5-2","max_tokens":128000,"provider_specific_entry":{"us":1.1},"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_output_config":true,"supports_forced_tool_choice":false,"supports_web_search":true,"supports_safeguards":true},"vertex_ai/claude-opus-latest":{"alias_of":"vertex_ai/claude-opus-5-5","base_model":"claude-opus"},"vertex_ai/deep-research-preview-04-2026":{"mode":"chat","source":"https://cloud.google.com/model-garden/pricing","web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"deep-research"},"vertex_ai/lyria-002":{"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing#lyria","supported_audio_formats":["wav"],"supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":true,"vertex_ai_audio_api":"lyria_predict","provider":"vertex_ai","base_model":"lyria-002"},"vertex_ai/lyria-3-clip-preview":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing#lyria","supported_audio_formats":["mp3"],"supported_endpoints":["/v1beta/interactions","/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supported_regions":["global"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"vertex_ai_audio_api":"lyria_interactions","provider":"vertex_ai","base_model":"lyria-3-clip"},"vertex_ai/lyria-3-pro-preview":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing#lyria","supported_audio_formats":["mp3","wav"],"supported_endpoints":["/v1beta/interactions","/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supported_regions":["global"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"vertex_ai_audio_api":"lyria_interactions","provider":"vertex_ai","base_model":"lyria-3-pro"},"vertex_ai/veo-3.1-lite-generate-001":{"max_input_tokens":1024,"mode":"video_generation","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing#veo","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"vertex_ai","base_model":"veo-3.1-lite-generate"},"voyage/rerank-1":{"max_input_tokens":8000,"max_output_tokens":8000,"mode":"rerank","source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","base_model":"rerank-1"},"voyage/rerank-lite-1":{"max_input_tokens":4000,"max_output_tokens":4000,"mode":"rerank","source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","base_model":"rerank-lite-1"},"voyage/rerank-3":{"max_input_tokens":32000,"max_output_tokens":32000,"mode":"rerank","source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","base_model":"rerank-3"},"voyage/rerank-3-lite":{"max_input_tokens":32000,"max_output_tokens":32000,"mode":"rerank","source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","base_model":"rerank-3-lite"},"voyage/voyage-large-2-instruct":{"max_input_tokens":16000,"mode":"embedding","output_vector_size":1024,"source":"https://docs.voyageai.com/docs/pricing","provider":"voyage","base_model":"voyage-large-2-instruct"},"xai/grok-latest":{"alias_of":"xai/grok-4.7","base_model":"grok"},"zai/glm-latest":{"alias_of":"zai/glm-5.3","base_model":"glm"},"zai/glm-flash-latest":{"alias_of":"zai/glm-5.3-flash","base_model":"glm-flash"},"zai/glm-5v-turbo":{"max_input_tokens":200000,"max_output_tokens":128000,"mode":"chat","source":"https://vercel.com/ai-gateway/models/glm-5v-turbo","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"zai","base_model":"glm-5v-turbo"},"vertex_ai/semantic-ranker-default":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-default"},"vertex_ai/semantic-ranker-default-002":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-default-002"},"vertex_ai/semantic-ranker-default-003":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-default-003"},"vertex_ai/semantic-ranker-default-004":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-default-004"},"vertex_ai/semantic-ranker-default@latest":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-default"},"vertex_ai/semantic-ranker-fast":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-fast"},"vertex_ai/semantic-ranker-fast-004":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-fast-004"},"vertex_ai/semantic-ranker-fast@latest":{"mode":"rerank","source":"https://cloud.google.com/generative-ai-app-builder/pricing","provider":"vertex_ai","base_model":"semantic-ranker-fast"},"runwayml/gen4.5":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"metadata":{"comment":"12 credits per second @ $0.01 per credit = $0.12 per second. The API uses ratio values like 1280:720; converted here to WxH resolution strings."},"provider":"runwayml","base_model":"gen4.5","supported_resolutions":["1280x720","720x1280","1104x832","960x960","832x1104"]},"runwayml/aleph2":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","video"],"supported_output_modalities":["video"],"metadata":{"comment":"28 credits per second @ $0.01 per credit = $0.28 per second; 56 credit minimum per task not modeled"},"provider":"runwayml","base_model":"aleph2"},"runwayml/seedance2":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"36 credits per second @ $0.01 per credit = $0.36 per second (480p/720p); 40 credits/s = $0.40 (1080p); 150 credits/s = $1.50 (4K). No credit minimum. Supports text-to-video, image-to-video, and video-to-video at the same per-second rate."},"provider":"runwayml","base_model":"seedance2"},"runwayml/seedance2_fast":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"11 credits per second at 480p, 24 at 720p @ $0.01 per credit; 44 credit minimum per generation not modeled"},"provider":"runwayml","base_model":"seedance2-fast"},"runwayml/seedance2_mini":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"7 credits per second at 480p, 15 at 720p @ $0.01 per credit; 28 credit minimum per generation not modeled"},"provider":"runwayml","base_model":"seedance2-mini"},"runwayml/seedance2_5":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"Output: 20/30/68/110 credits per second at 480p/720p/1080p/4K @ $0.01 per credit; input video billed additionally at 10/15/34/57 credits per input second and the 80 credit minimum per task are not modeled"},"provider":"runwayml","base_model":"seedance2-5"},"runwayml/seedance-latest":{"alias_of":"runwayml/seedance2_5","base_model":"seedance"},"runwayml/hailuo3":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"10 credits per second at 768P, 15 at 2K (mapped to the 1080p tier) @ $0.01 per credit; 2 credits per reference image not modeled"},"provider":"runwayml","base_model":"hailuo3"},"runwayml/gemini_omni_flash":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"metadata":{"comment":"10 credits per second @ $0.01 per credit = $0.10 per second"},"provider":"runwayml","base_model":"gemini-omni-flash"},"runwayml/veo3.1":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"metadata":{"comment":"40 credits per second with audio, 20 without @ $0.01 per credit; priced at the with-audio rate"},"provider":"runwayml","base_model":"veo3.1"},"runwayml/veo3.1_fast":{"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"metadata":{"comment":"15 credits per second with audio, 10 without @ $0.01 per credit; priced at the with-audio rate"},"provider":"runwayml","base_model":"veo3.1-fast"},"scaleway/openai/whisper-large-v3":{"mode":"audio_transcription","provider":"scaleway","base_model":"whisper-large-v3"},"gpt-realtime-whisper":{"mode":"audio_transcription","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"openai","base_model":"gpt-realtime-whisper","web_search_billing_unit":"per_request"},"gemini/gemini-3.1-flash-live-preview":{"successor_model":"gemini/gemini-3.8-live","max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"rpm":10,"gemini_audio_only_live":true,"supports_response_schema":false,"supports_reasoning":true,"provider":"gemini","base_model":"gemini-3.1-flash-live"},"gemini/gemini-3.1-flash-tts-preview":{"deprecation_date":"2026-12-18","deprecation_source":"https://ai.google.dev/gemini-api/docs/deprecations","successor_model":"gemini/gemini-3.8-flash-tts","max_input_tokens":8192,"max_output_tokens":16384,"mode":"audio_speech","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"tpm":4000000,"rpm":10,"supports_function_calling":false,"supports_response_schema":false,"supports_web_search":false,"provider":"gemini","base_model":"gemini-3.1-flash-tts"},"gemini/gemini-3.8-flash-tts":{"max_input_tokens":8192,"max_output_tokens":16384,"mode":"audio_speech","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"supports_prompt_caching":false,"provider":"gemini","base_model":"gemini-3.8-flash-tts"},"gemini/gemini-3.8-flash-lite-tts":{"max_input_tokens":8192,"max_output_tokens":16384,"mode":"audio_speech","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/audio/speech"],"supports_prompt_caching":false,"provider":"gemini","base_model":"gemini-3.8-flash-lite-tts"},"duckduckgo/search":{"mode":"search","metadata":{"notes":"DuckDuckGo Instant Answer API is free and does not require an API key."},"provider":"duckduckgo","base_model":"search"},"us.openai.gpt-daybreak-blue-5.6-sol":{"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-daybreak-blue-56-sol.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","base_model":"gpt-daybreak-blue-5.6-sol"},"global.openai.gpt-daybreak-blue-5.6-sol":{"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-daybreak-blue-56-sol.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","base_model":"gpt-daybreak-blue-5.6-sol"},"us.openai.gpt-5.6-cyber":{"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-cyber.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","base_model":"gpt-5.6-cyber"},"global.openai.gpt-5.6-cyber":{"supports_bedrock_runtime_chat_completions_response_format":true,"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-cyber.html","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_tool_choice":true,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"provider":"bedrock","base_model":"gpt-5.6-cyber"},"us.openai.gpt-5.4":{"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":131072,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_minimal_reasoning_effort":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/responses"],"provider":"bedrock","base_model":"gpt-5.4","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-west-1","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","us-gov-east-1"],"max_tokens":128000},"global.openai.gpt-5.4":{"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":131072,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_minimal_reasoning_effort":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/responses"],"provider":"bedrock","base_model":"gpt-5.4","supports_prompt_caching":true,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supported_regions":["us-east-1","us-east-2","us-west-2","us-gov-west-1","us-west-1","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","us-gov-east-1"],"max_tokens":128000},"us.openai.gpt-5.5":{"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":131072,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_minimal_reasoning_effort":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/responses"],"provider":"bedrock","base_model":"gpt-5.5","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supports_prompt_caching":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000},"global.openai.gpt-5.5":{"supports_bedrock_converse_tool_result_images":false,"max_input_tokens":922000,"max_output_tokens":131072,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_minimal_reasoning_effort":false,"supports_tool_choice":false,"supports_reasoning":true,"supports_xhigh_reasoning_effort":true,"supports_vision":true,"supports_sampling_params":false,"supported_endpoints":["/v1/responses"],"provider":"bedrock","base_model":"gpt-5.5","supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supports_prompt_caching":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","mx-central-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1"],"max_tokens":128000},"bedrock_mantle/gpt-astra-latest":{"alias_of":"bedrock_mantle/openai.gpt-6-astra","base_model":"gpt-astra"},"bedrock_mantle/gpt-luna-latest":{"alias_of":"bedrock_mantle/openai.gpt-6-luna","base_model":"gpt-luna"},"bedrock/gpt-astra-latest":{"alias_of":"global.openai.gpt-6-astra","base_model":"gpt-astra"},"bedrock/gpt-luna-latest":{"alias_of":"global.openai.gpt-6-luna","base_model":"gpt-luna"},"bedrock_mantle/claude-opus-latest":{"alias_of":"bedrock_mantle/anthropic.claude-opus-5-5","base_model":"claude-opus"},"bedrock_mantle/claude-sonnet-latest":{"alias_of":"bedrock_mantle/anthropic.claude-sonnet-5-5","base_model":"claude-sonnet"},"volcengine/doubao-seed-2-1-pro-260915":{"max_input_tokens":1024000,"max_output_tokens":256000,"mode":"chat","source":"https://www.volcengine.com/docs/82379/1544106","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"provider":"volcengine","base_model":"doubao-seed-2-1-pro"},"volcengine/doubao-seed-2-1-lite-260915":{"max_input_tokens":1024000,"max_output_tokens":256000,"mode":"chat","source":"https://www.volcengine.com/docs/82379/1544106","supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":false,"supports_vision":true,"provider":"volcengine","base_model":"doubao-seed-2-1-lite"},"volcengine/doubao-seedance-2-5-260628":{"mode":"video_generation","source":"https://www.volcengine.com/docs/82379/1544106","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"volcengine_video_frames":{"480p":{"16:9":"854x480","4:3":"752x560","1:1":"640x640","3:4":"560x752","9:16":"480x854","21:9":"992x432"},"720p":{"16:9":"1280x720","4:3":"1112x834","1:1":"960x960","3:4":"834x1112","9:16":"720x1280","21:9":"1470x630"}}},"metadata":{"comment":"Ark bills per output token, not per second: tokens = width * height * 24 * seconds / 1024, so a per-second rate per resolution tier is exact for the 16:9 geometry Ark publishes (1280x720 is 21600 tokens/s). List 70 CNY per million tokens at 480p and 720p and 77 at 1080p without a video input, matching Ark's per-resolution table and its 1080p 16:9 example of 3.74 CNY/s, converted at 7.0 CNY/USD like the other volcengine entries. Accepts up to 50 reference assets (30 images, 10 videos, 10 audios) and clips of 4-30s; per-asset billing beyond the free allowance is not encoded. Requests with a video input bill at a lower rate per token and are rejected, since their cost is not known at create time. provider_specific_entry.volcengine_video_frames is the frame Ark renders per resolution tier and ratio, verified with ffprobe at 480p for every ratio and at 720p for 16:9 and 4:3; the other 720p ratios are from the Seedance 2.0 model card."},"provider":"volcengine","base_model":"doubao-seedance-2-5"},"volcengine/doubao-seedance-2-0-260128":{"mode":"video_generation","source":"https://www.volcengine.com/docs/82379/1544106","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"volcengine_video_frames":{"480p":{"16:9":"864x496","4:3":"752x560","1:1":"640x640","3:4":"560x752","9:16":"496x864","21:9":"992x432"},"720p":{"16:9":"1280x720","4:3":"1112x834","1:1":"960x960","3:4":"834x1112","9:16":"720x1280","21:9":"1470x630"}}},"metadata":{"comment":"Ark bills per output token, not per second: tokens = width * height * 24 * seconds / 1024, so a per-second rate per resolution tier is exact for the 16:9 geometry Ark publishes (1280x720 is 21600 tokens/s; the 480p frame is 864x496, 10044 tokens/s, not the tier's nominal 854x480). List 46 CNY per million tokens at 480p and 720p, 51 at 1080p, and 26 at 4k without a video input, matching Ark's per-resolution table and its 1080p and 4k 16:9 examples of 2.48 and 5.05 CNY/s, converted at 7.0 CNY/USD like the other volcengine entries. Clips of 4-15s; 4k is exclusive to this tier. Requests with a video input bill at a lower rate per token and are rejected, since their cost is not known at create time. provider_specific_entry.volcengine_video_frames is the frame Ark renders per resolution tier and ratio, verified with ffprobe at 480p for every ratio and at 720p for 16:9 and 4:3; the other 720p ratios are from the Seedance 2.0 model card."},"provider":"volcengine","base_model":"doubao-seedance-2-0"},"volcengine/doubao-seedance-2-0-fast-260128":{"mode":"video_generation","source":"https://www.volcengine.com/docs/82379/1544106","supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"volcengine_video_frames":{"480p":{"16:9":"864x496","4:3":"752x560","1:1":"640x640","3:4":"560x752","9:16":"496x864","21:9":"992x432"},"720p":{"16:9":"1280x720","4:3":"1112x834","1:1":"960x960","3:4":"834x1112","9:16":"720x1280","21:9":"1470x630"}}},"metadata":{"comment":"Ark bills per output token, not per second: tokens = width * height * 24 * seconds / 1024, so a per-second rate per resolution tier is exact for the 16:9 geometry Ark publishes (1280x720 is 21600 tokens/s; the 480p frame is 864x496, 10044 tokens/s, not the tier's nominal 854x480). List 37 CNY per million tokens without a video input, converted at 7.0 CNY/USD like the other volcengine entries. Clips of 4-15s at 480p or 720p. Requests with a video input bill at a lower rate per token and are rejected, since their cost is not known at create time. provider_specific_entry.volcengine_video_frames is the frame Ark renders per resolution tier and ratio, verified with ffprobe at 480p for every ratio and at 720p for 16:9 and 4:3; the other 720p ratios are from the Seedance 2.0 model card."},"provider":"volcengine","base_model":"doubao-seedance-2-0-fast"},"volcengine/doubao-seedance-2-0-mini-260615":{"mode":"video_generation","source":"https://www.volcengine.com/docs/82379/1544106","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supported_endpoints":["/v1/videos"],"provider_specific_entry":{"volcengine_video_frames":{"480p":{"16:9":"864x496","4:3":"752x560","1:1":"640x640","3:4":"560x752","9:16":"496x864","21:9":"992x432"},"720p":{"16:9":"1280x720","4:3":"1112x834","1:1":"960x960","3:4":"834x1112","9:16":"720x1280","21:9":"1470x630"}}},"metadata":{"comment":"Ark bills per output token, not per second: tokens = width * height * 24 * seconds / 1024, so a per-second rate per resolution tier is exact for the 16:9 geometry Ark publishes (1280x720 is 21600 tokens/s; the 480p frame is 864x496, 10044 tokens/s, not the tier's nominal 854x480). List 23 CNY per million tokens without a video input, converted at 7.0 CNY/USD like the other volcengine entries. Clips of 4-15s at 480p or 720p. Only text and image inputs are declared; the mini tier's reference video and audio caps were not verified. Requests with a video input bill at a lower rate per token and are rejected, since their cost is not known at create time. provider_specific_entry.volcengine_video_frames is the frame Ark renders per resolution tier and ratio, verified with ffprobe at 480p for every ratio and at 720p for 16:9 and 4:3; the other 720p ratios are from the Seedance 2.0 model card."},"provider":"volcengine","base_model":"doubao-seedance-2-0-mini"},"soniox/stt-async-v4":{"max_output_tokens":8000,"max_input_tokens":8000,"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"supports_audio_input":true,"deprecation_date":"2026-06-30","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-async-v5","provider":"soniox","base_model":"stt-async-v4","is_deprecated":true},"soniox/stt-async-v5":{"max_output_tokens":8000,"max_input_tokens":8000,"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"supports_audio_input":true,"deprecation_date":"2026-12-15","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-async-v6","provider":"soniox","base_model":"stt-async-v5"},"soniox/stt-async-v6":{"max_output_tokens":8000,"max_input_tokens":8000,"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/audio/transcriptions"],"supports_audio_input":true,"provider":"soniox","base_model":"stt-async-v6"},"soniox/stt-rt-v5":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"deprecation_date":"2026-12-15","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-rt-v6","provider":"soniox","base_model":"stt-rt-v5"},"soniox/stt-rt-v6":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"soniox","base_model":"stt-rt-v6"},"soniox/stt-rt-preview":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"soniox","base_model":"stt-rt"},"soniox/stt-rt-preview-v2":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"deprecation_date":"2025-10-31","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-rt-v3","provider":"soniox","base_model":"stt-rt-v2","is_deprecated":true},"soniox/stt-rt-v3":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"deprecation_date":"2026-02-28","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-rt-v4","provider":"soniox","base_model":"stt-rt-v3","is_deprecated":true},"soniox/stt-rt-v3-preview":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"soniox","base_model":"stt-rt-v3"},"soniox/stt-rt-v4":{"metadata":{"calculation":"$0.12/3600 seconds = $0.00003333 per second","original_pricing_per_hour":0.12},"mode":"audio_transcription","source":"https://soniox.com/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"deprecation_date":"2026-06-30","deprecation_source":"https://docs.soniox.com/stt/models","successor_model":"soniox/stt-rt-v5","provider":"soniox","base_model":"stt-rt-v4","is_deprecated":true},"streamlake/deepseek-v4-flash-0731":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4-flash"},"streamlake/DeepSeek-V4-Pro-0813":{"max_input_tokens":1048566,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4-pro"},"streamlake/DeepSeek-V3":{"max_input_tokens":56000,"max_output_tokens":8000,"mode":"chat","supports_function_calling":true,"supports_reasoning":false,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v3"},"streamlake/GLM-5.3":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_assistant_prefill":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"glm-5.3"},"streamlake/GLM-5.3-Flash":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_assistant_prefill":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"glm-5.3-flash"},"streamlake/Kimi-K2-Instruct-0905":{"max_input_tokens":224000,"max_output_tokens":32000,"mode":"chat","supports_function_calling":true,"supports_reasoning":false,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"kimi-k2-instruct"},"streamlake/deepseek-r1":{"max_input_tokens":98304,"max_output_tokens":16000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-r1"},"streamlake/deepseek-v3.1":{"max_input_tokens":98304,"max_output_tokens":32000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek"},"streamlake/deepseek-v4-flash":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4-flash"},"streamlake/deepseek-v4-pro":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4-pro"},"streamlake/deepseek-v4.1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4.1-flash"},"streamlake/deepseek-v4.1-flash-260910":{"max_input_tokens":1048568,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"deepseek-v4.1-flash"},"streamlake/glm-5":{"max_input_tokens":202752,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"glm-5"},"streamlake/glm-5.1":{"max_input_tokens":202752,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"glm-5.1"},"streamlake/glm-5.2":{"max_input_tokens":1048566,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"glm-5.2"},"streamlake/kimi-k2.6":{"max_input_tokens":262144,"max_output_tokens":256000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_reasoning":false,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"kimi-k2.6"},"streamlake/kimi-k2.7-code":{"max_input_tokens":262144,"max_output_tokens":32000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"kimi-k2.7-code"},"streamlake/mimo-v2.6-flash":{"mode":"chat","supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"mimo-v2.6-flash"},"streamlake/mimo-v2.6-pro":{"mode":"chat","supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"mimo-v2.6-pro"},"streamlake/minimax-m2.5":{"max_input_tokens":196608,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"minimax-m2.5"},"streamlake/minimax-m3":{"max_input_tokens":1000000,"max_output_tokens":512000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"minimax-m3"},"streamlake/qwen3.5-397b-a17b":{"max_input_tokens":258048,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"qwen3.5-397b-a17b"},"streamlake/qwen3.6-27b":{"max_input_tokens":229376,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_reasoning":false,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"qwen3.6-27b"},"streamlake/qwen3.6-35b-a3b":{"max_input_tokens":229376,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_reasoning":false,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"qwen3.6-35b-a3b"},"streamlake/qwen3.8-2.4t-a95b":{"max_input_tokens":983616,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"qwen3.8-2.4t-a95b"},"streamlake/qwen3.8-27b":{"max_input_tokens":983616,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://www.streamlake.ai/document/DOC/mgrnm4xm362hvp5wyce","provider":"streamlake","base_model":"qwen3.8-27b"},"deepseek/deepseek-flash-latest":{"alias_of":"deepseek/deepseek-v4-flash","base_model":"deepseek-flash"},"deepseek/deepseek-pro-latest":{"alias_of":"deepseek/deepseek-v4-pro","base_model":"deepseek-pro"},"tencent/deepseek-v4-pro-0813":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","off_peak_pricing":{"cache_read_input_token_cost":2.2e-8,"input_cost_per_token":6.6e-7,"output_cost_per_token":0.00000198,"weekday_timezone":"Asia/Shanghai","windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":false,"provider":"tencent","base_model":"deepseek-v4-pro"},"tencent/deepseek-v4.1-flash":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","off_peak_pricing":{"cache_read_input_token_cost":3e-9,"input_cost_per_token":1.5e-7,"output_cost_per_token":6e-7,"weekday_timezone":"Asia/Shanghai","windows":[{"hours_utc":["00:00-01:00","04:00-06:00","10:00-00:00"],"weekdays":[1,2,3,4,5]},{"hours_utc":"00:00-00:00","weekdays":[6,7]}]},"source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"provider":"tencent","base_model":"deepseek-v4.1-flash"},"tencent/glm-5.3-flashx":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"tencent","base_model":"glm-5.3-flashx"},"tencent/glm-5.3-flash":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"tencent","base_model":"glm-5.3-flash"},"tencent/glm-5.3":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"glm-5.3"},"tencent/glm-5.2":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"glm-5.2"},"tencent/kimi-k3":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"kimi-k3"},"tencent/kimi-k2.8-preview":{"max_input_tokens":1000000,"max_output_tokens":1000000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"kimi-k2.8"},"tencent/kimi-k2.7-code-highspeed":{"max_input_tokens":256000,"max_output_tokens":256000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"kimi-k2.7-code-highspeed"},"tencent/kimi-k2.7-code":{"max_input_tokens":256000,"max_output_tokens":256000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"tencent","base_model":"kimi-k2.7-code"},"tencent/hy4":{"comment":"These are the 50% promotional rates. Update on 2026-12-31 when the promotion ends at 23:59:59 Beijing time. List rates per 1M tokens are $0.834 input, $2.501 output and $0.042 cached input.","max_input_tokens":960000,"max_output_tokens":64000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"tencent","base_model":"hy4"},"tencent/hy4-preview":{"comment":"These are the 50% promotional rates. Update on 2026-12-31 when the promotion ends at 23:59:59 Beijing time. List rates per 1M tokens are $0.834 input, $2.501 output and $0.042 cached input.","max_input_tokens":960000,"max_output_tokens":64000,"mode":"chat","source":"https://www.tencentcloud.com/document/product/1300/78937","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"provider":"tencent","base_model":"hy4"},"tencent/seedream-image-v5.0-flash":{"mode":"image_generation","source":"https://cloud.tencent.com/document/product/1823/130055","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"tencent","base_model":"seedream-image-v5.0-flash"},"tencent/seedream-image-v5.0-lite":{"mode":"image_generation","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"tencent","base_model":"seedream-image-v5.0-lite"},"tencent/seedream-image-v5.0-pro":{"mode":"image_generation","source":"https://www.tencentcloud.com/document/product/1300/78937","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"tencent","base_model":"seedream-image-v5.0-pro"},"gpt-transcribe":{"mode":"audio_transcription","provider_specific_entry":{"transcription_json_or_text_only":1,"transcription_usage_duration":1},"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/audio/transcriptions","/v1/realtime/transcription_sessions"],"supported_modalities":["audio","text"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_native_streaming":true,"provider":"openai","base_model":"gpt-transcribe","web_search_billing_unit":"per_request"},"gpt-live-transcribe":{"mode":"audio_transcription","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_native_streaming":true,"provider":"openai","base_model":"gpt-live-transcribe","web_search_billing_unit":"per_request"},"gpt-live-1":{"mode":"realtime","source":"https://developers.openai.com/api/docs/pricing","supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"openai","base_model":"gpt-live-1","web_search_billing_unit":"per_request"},"gpt-realtime-translate":{"max_input_tokens":16000,"max_output_tokens":2000,"mode":"realtime","source":"https://developers.openai.com/api/docs/pricing","supported_modalities":["audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_native_streaming":true,"provider":"openai","base_model":"gpt-realtime-translate","web_search_billing_unit":"per_request"},"claude-mythos-5-2":{"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_always_on":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":512,"supports_native_structured_output":true,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_fast_mode":true,"supports_speed":true,"provider":"anthropic","base_model":"claude-mythos-5-2","web_search_billing_unit":"per_request","max_tokens":1000000,"inference_geo_us_multiplier":1.1,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"supports_reasoning_with_tool_calls":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_mid_conversation_system_messages":true,"supports_web_search":true,"supports_safeguards":true,"supports_service_tier":true,"supports_native_effort":true,"supports_reasoning_effort":true,"supports_dynamic_reasoning_budget":false,"supports_response_schema_with_tools":true,"service_tiers":["auto","standard_only","fast"]},"mistral/mistral-moderation-2603":{"max_input_tokens":131072,"mode":"moderation","source":"https://docs.mistral.ai/models/model-cards/mistral-moderation-26-03","provider":"mistral","base_model":"mistral-moderation"},"mistral/voxtral-mini-2602":{"deprecation_date":"2027-04-30","mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini"},"mistral/voxtral-mini-transcribe-realtime-2602":{"deprecation_date":"2027-04-30","mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-realtime-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-transcribe-realtime"},"mistral/voxtral-mini-tts-2603":{"deprecation_date":"2027-04-30","mode":"audio_speech","source":"https://docs.mistral.ai/models/model-cards/voxtral-tts-26-03","supported_endpoints":["/v1/audio/speech"],"supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_output":true,"provider":"mistral","base_model":"voxtral-mini-tts"},"fallback_generalizations":{"rules":[{"name":"anthropic-claude-adaptive-thinking","pattern":"(?:opus|sonnet|haiku)[-._](?:4[-._](?:[6-9]|[1-9]\\d)(?!\\d)|(?:[5-9]|[1-9]\\d{1,})[-._]\\d{1,2}(?!\\d))","description":"Claude opus/sonnet/haiku at version 4.6 or higher: 4.6 through 4.99, then any 5.x, 6.x or later major. The minor is capped at two digits so an 8-digit date suffix such as claude-opus-4-20250514 is never read as a >= 4.6 minor. Turns on adaptive thinking for new families with no code change.","extends":"anthropic-claude","model_info":{"supports_adaptive_thinking":true}},{"name":"anthropic-claude","pattern":"^claude-[a-z]+-\\d+[-.]\\d+(?:-\\d{8})?$","description":"Any Claude family-major-minor id, optionally with an 8-digit date suffix, anchored to the whole name. Version-neutral fallback that gives an unmapped Claude provider routing and baseline capabilities; it carries no pricing, so cost stays on the standard unpriced behavior rather than a guessed number.","model_info":{"mode":"chat","max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_tool_choice":true,"supports_assistant_prefill":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_reasoning":true,"supports_pdf_input":true,"supports_system_messages":true}}],"base_model":"fallback-generalizations"},"gemini/gemini-3.5-live-translate-preview":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","rpm":10,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["audio","text"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":false,"supports_response_schema":false,"supports_web_search":false,"tpm":250000,"provider":"gemini","base_model":"gemini-3.5-live-translate"},"gemini/gemini-3.5-transcribe":{"max_input_tokens":98304,"max_output_tokens":32768,"mode":"audio_transcription","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"tpm":800000,"rpm":2000,"supports_function_calling":false,"provider":"gemini","base_model":"gemini-3.5-transcribe"},"gemini/gemini-3.5-transcribe-live":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"audio_transcription","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"tpm":250000,"rpm":10,"supports_function_calling":false,"provider":"gemini","base_model":"gemini-3.5-transcribe-live"},"vertex_ai/gemini-3.5-transcribe-preview":{"mode":"audio_transcription","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"vertex_ai","base_model":"gemini-3.5-transcribe"},"vertex_ai/gemini-3.5-transcribe-live-preview":{"mode":"audio_transcription","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"vertex_ai","base_model":"gemini-3.5-transcribe-live"},"vertex_ai/gemini-3.5-live-translate-preview":{"mode":"realtime","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["audio","text"],"supports_audio_input":true,"supports_audio_output":true,"provider":"vertex_ai","base_model":"gemini-3.5-live-translate"},"fireworks_ai/accounts/fireworks/routers/deepseek-v4p1-flash-us":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/deepseek-v4.1-flash-us"},"fireworks_ai/accounts/fireworks/routers/kimi-latest":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/kimi"},"fireworks_ai/kimi-latest":{"alias_of":"fireworks_ai/accounts/fireworks/models/kimi-k3","base_model":"kimi"},"fireworks_ai/deepseek-v4p1-flash-us":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"deepseek-v4.1-flash-us"},"fireworks_ai/accounts/fireworks/models/muse-spark-open":{"max_input_tokens":1048576,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"muse-spark-open","supports_prompt_caching":true,"max_tokens":1048576},"fireworks_ai/accounts/fireworks/routers/qwen-max-latest":{"max_input_tokens":262144,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/qwen-max"},"fireworks_ai/accounts/fireworks/routers/kimi-fast-latest":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/kimi-fast"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1048576,"max_output_tokens":131072,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash","provider":"deepinfra","base_model":"deepseek-v4.1-flash"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"max_input_tokens":1048576,"max_output_tokens":131072,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"source":"https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash-1017","provider":"deepinfra","base_model":"deepseek-v4.1-flash"},"deepinfra/zai-org/GLM-5.3-Flash":{"max_output_tokens":1048576,"max_input_tokens":1048576,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"supports_video_input":true,"source":"https://deepinfra.com/zai-org/GLM-5.3-Flash","provider":"deepinfra","base_model":"glm-5.3-flash"},"deepinfra/XiaomiMiMo/MiMo-V2.6-Flash":{"max_input_tokens":1048576,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"supports_video_input":true,"supports_audio_input":true,"source":"https://deepinfra.com/XiaomiMiMo/MiMo-V2.6-Flash","provider":"deepinfra","base_model":"mimo-v2.6-flash"},"deepinfra/nvidia/NVIDIA-Nemotron-3.5-Super-VL":{"max_input_tokens":262144,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":true,"supports_video_input":true,"source":"https://deepinfra.com/nvidia/NVIDIA-Nemotron-3.5-Super-VL","provider":"deepinfra","base_model":"nvidia-nemotron-3.5-super-vl"},"deepinfra/tencent/Hy4-preview":{"max_output_tokens":131072,"max_input_tokens":1048576,"supports_prompt_caching":true,"mode":"chat","supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_reasoning":true,"supports_vision":false,"source":"https://deepinfra.com/tencent/Hy4-preview","provider":"deepinfra","base_model":"hy4"},"xai/grok-imagine-image":{"mode":"image_generation","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image","max_input_tokens":16000,"max_tokens":16000,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false},"xai/grok-imagine-image-2026-03-02":{"mode":"image_generation","source":"https://docs.x.ai/docs/models","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image","max_input_tokens":16000,"max_tokens":16000,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false},"xai/grok-imagine-image-quality":{"deprecation_date":"2026-11-02","deprecation_source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","mode":"image_generation","source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","successor_model":"xai/grok-imagine-image-2.0","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image-quality","is_deprecated":true},"xai/grok-imagine-image-quality-20260403":{"deprecation_date":"2026-11-02","deprecation_source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","mode":"image_generation","source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","successor_model":"xai/grok-imagine-image-2.0","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image-quality","is_deprecated":true},"xai/grok-imagine-image-quality-latest":{"deprecation_date":"2026-11-02","deprecation_source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","mode":"image_generation","source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","successor_model":"xai/grok-imagine-image-2.0","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image-quality","is_deprecated":true},"xai/grok-imagine-image-pro":{"deprecation_date":"2026-05-15","deprecation_source":"https://docs.x.ai/developers/migration/may-15-retirement","mode":"image_generation","source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","successor_model":"xai/grok-imagine-image-2.0","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image-pro","is_deprecated":true},"xai/grok-imagine-image-2.0":{"mode":"image_generation","source":"https://docs.x.ai/developers/migration/imagine-image-quality-nov-2","supported_endpoints":["/v1/images/generations"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"xai","base_model":"grok-imagine-image-2.0","max_input_tokens":64000,"max_tokens":64000,"supports_vision":true,"supports_audio_input":false,"supports_audio_output":false},"xai/grok-imagine-video":{"mode":"video_generation","source":"https://docs.x.ai/docs/models/grok-imagine-video","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"xai","base_model":"grok-imagine-video"},"xai/grok-imagine-video-1.5":{"mode":"video_generation","source":"https://docs.x.ai/docs/models/grok-imagine-video-1.5","supported_modalities":["text","image","audio"],"supported_output_modalities":["video"],"provider":"xai","base_model":"grok-imagine-video-1.5"},"xai/grok-imagine-video-1.5-lite":{"mode":"video_generation","source":"https://docs.x.ai/developers/pricing","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"xai","base_model":"grok-imagine-video-1.5-lite"},"xai/grok-imagine-video-1.5-2026-05-30":{"mode":"video_generation","source":"https://docs.x.ai/docs/models/grok-imagine-video-1.5","supported_modalities":["text","image","audio"],"supported_output_modalities":["video"],"provider":"xai","base_model":"grok-imagine-video-1.5"},"xai/grok-imagine-video-1.5-preview":{"mode":"video_generation","source":"https://docs.x.ai/docs/models/grok-imagine-video-1.5","supported_modalities":["text","image","audio"],"supported_output_modalities":["video"],"provider":"xai","base_model":"grok-imagine-video-1.5"},"xai/grok-voice-transcribe-1.0":{"deprecation_date":"2026-10-02","deprecation_source":"https://docs.x.ai/docs/release-notes","successor_model":"xai/grok-voice-transcribe-2.0","metadata":{"calculation":"$0.10/3600 seconds = $0.00002778 per second","original_pricing_per_hour":0.1},"mode":"audio_transcription","source":"https://docs.x.ai/developers/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"xai","base_model":"grok-voice-transcribe-1.0","is_deprecated":true},"xai/grok-voice-transcribe-2.0":{"metadata":{"calculation":"$0.10/3600 seconds = $0.00002778 per second","original_pricing_per_hour":0.1},"mode":"audio_transcription","source":"https://docs.x.ai/developers/pricing","supported_endpoints":["/v1/audio/transcriptions"],"provider":"xai","base_model":"grok-voice-transcribe-2.0"},"mistral/mistral-ocr-3":{"mode":"ocr","supported_endpoints":["/v1/ocr","/v1/batch"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","base_model":"mistral-ocr-3"},"mistral/mistral-ocr-3-0":{"mode":"ocr","supported_endpoints":["/v1/ocr","/v1/batch"],"source":"https://mistral.ai/pricing#api-pricing","provider":"mistral","base_model":"mistral-ocr-3-0"},"mistral/mistral-ocr-4":{"mode":"ocr","source":"https://docs.mistral.ai/models/model-cards/ocr-4-1","supported_endpoints":["/v1/ocr","/v1/batch"],"provider":"mistral","base_model":"mistral-ocr-4"},"mistral/voxtral-mini-latest":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini"},"mistral/voxtral-mini-realtime-2602":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-realtime-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-realtime"},"mistral/voxtral-mini-realtime-latest":{"mode":"audio_transcription","source":"https://docs.mistral.ai/models/model-cards/voxtral-mini-transcribe-realtime-26-02","supported_endpoints":["/v1/audio/transcriptions"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider":"mistral","base_model":"voxtral-mini-realtime"},"fireworks_ai/accounts/fireworks/routers/glm-latest":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm"},"fireworks_ai/accounts/fireworks/routers/glm-5p3-us":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm-5.3-us"},"fireworks_ai/glm-5p3-us":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"glm-5.3-us"},"fireworks_ai/accounts/fireworks/routers/glm-fast-latest":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm-fast"},"fireworks_ai/accounts/fireworks/routers/auto":{"mode":"chat","source":"https://docs.fireworks.ai/nexus/firerouter#example-router-ids","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/auto"},"fireworks_ai/accounts/fireworks/routers/auto-instant":{"mode":"chat","source":"https://docs.fireworks.ai/nexus/firerouter#example-router-ids","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/auto-instant"},"fireworks_ai/accounts/fireworks/routers/firerouter":{"mode":"chat","source":"https://docs.fireworks.ai/nexus/firerouter#example-router-ids","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/firerouter"},"fireworks_ai/accounts/fireworks/routers/glm-flash-latest":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm-flash"},"fireworks_ai/accounts/fireworks/routers/glm-5p3-flash-us":{"deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://docs.fireworks.ai/changelog.md","successor_model":"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash-1017-us","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm-5.3-flash-us","is_deprecated":true},"fireworks_ai/glm-5p3-flash-us":{"deprecation_date":"2026-11-07","deprecation_source":"https://docs.fireworks.ai/changelog.md","max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://docs.fireworks.ai/changelog.md","successor_model":"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash-1017-us","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"glm-5.3-flash-us","is_deprecated":true},"zerogpu/deepseek-v4.1-flash":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"deepseek-v4.1-flash"},"zerogpu/glm-5.3-flash":{"max_input_tokens":1048576,"max_output_tokens":32768,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"glm-5.3-flash"},"zerogpu/gpt-4.1-mini":{"max_input_tokens":1047576,"max_output_tokens":32768,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"gpt-4.1-mini"},"zerogpu/gpt-5.4-nano":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"gpt-5.4-nano"},"zerogpu/gpt-5.6-luna":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"gpt-5.6-luna"},"zerogpu/gpt-6-luna":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"zerogpu","base_model":"gpt-6-luna"},"zerogpu/gpt-oss-120b":{"max_input_tokens":128000,"max_output_tokens":32768,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"gpt-oss-120b"},"zerogpu/LFM2.5-1.2B-Instruct":{"max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","provider":"zerogpu","base_model":"lfm2.5-1.2b-instruct"},"zerogpu/LFM2.5-1.2B-Thinking":{"max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","provider":"zerogpu","base_model":"lfm2.5-1.2b-thinking"},"zerogpu/llama-3.1-8b-instruct-fast":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"llama-3.1-8b-instruct-fast"},"zerogpu/llama-3.2-3b-instruct-fast":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","provider":"zerogpu","base_model":"llama-3.2-3b-instruct-fast"},"zerogpu/qwen3-30b-a3b-fp8":{"max_input_tokens":32768,"max_output_tokens":16384,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"zerogpu","base_model":"qwen3-30b-a3b-fp8"},"zerogpu/qwen3.6-35b-a3b":{"max_input_tokens":262144,"max_output_tokens":16384,"mode":"chat","source":"https://docs.zerogpu.ai/docs/model-catalog","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"zerogpu","base_model":"qwen3.6-35b-a3b"},"together_ai/qwen-flash-latest":{"alias_of":"together_ai/Qwen/Qwen3.8-Flash","base_model":"qwen-flash"},"elevenlabs/scribe_v2":{"mode":"audio_transcription","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/audio/transcriptions"],"provider":"elevenlabs","base_model":"scribe-v2"},"azure/deepseek-v4.1-flash-1017":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"supports_prompt_caching":true,"supports_vision":false,"source":"https://ai.azure.com/catalog/models/DeepSeek-V4.1-Flash-1017","provider":"azure","base_model":"deepseek-v4.1-flash"},"azure/deepseek-v4.1-flash":{"max_input_tokens":1000000,"max_output_tokens":1048576,"mode":"chat","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"deprecation_date":"2026-12-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/deepseek-v4-1-flash-is-coming-to-microsoft-foundry/4556431","provider":"azure","base_model":"deepseek-v4.1-flash"},"azure/MAI-Image-2.6":{"max_input_tokens":32000,"mode":"image_generation","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","base_model":"mai-image-2.6"},"azure/MAI-Image-2.6-Flash":{"max_input_tokens":32000,"mode":"image_generation","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"provider":"azure","base_model":"mai-image-2.6-flash"},"azure/FW-DeepSeek-V4.1-Flash":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"fw-deepseek-v4.1-flash"},"azure/FW-DeepSeek-V4-Flash":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"fw-deepseek-v4-flash"},"azure/FW-GLM-5.3":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"fw-glm-5.3","supports_reasoning_summary_detailed":false,"supports_reasoning_summary_concise":false,"supports_reasoning_summary_auto":true,"supported_endpoints":["/v1/chat/completions"]},"azure/glm-latest":{"alias_of":"azure_ai/FW-GLM-5.3","base_model":"glm"},"azure/FW-GLM-5.3-Flash":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"fw-glm-5.3-flash","supports_reasoning_summary":false,"supported_endpoints":["/v1/chat/completions"]},"azure/glm-flash-latest":{"alias_of":"azure_ai/FW-GLM-5.3-Flash","base_model":"glm-flash"},"azure/FW-GPT-OSS-120B":{"deprecation_date":"2027-07-01","max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"azure","base_model":"fw-gpt-oss-120b"},"azure/Cohere-command-a-plus-05-2026":{"deprecation_date":"2026-10-16","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"coherelabs-command-a-plus-05-2026-fp81","max_input_tokens":128000,"max_output_tokens":64000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"command-a-plus","is_deprecated":true},"azure/mistral-medium-3-5":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_pdf_input":false,"supports_response_schema":true,"supports_vision":true,"provider":"azure","base_model":"mistral-medium-3-5"},"azure/deepseek-v3.1":{"deprecation_date":"2026-07-13","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","successor_model":"azure_ai/deepseek-v4-flash","mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supports_pdf_input":false,"provider":"azure","base_model":"deepseek","is_deprecated":true},"bedrock_mantle/zai.glm-5":{"max_input_tokens":200000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"supports_native_structured_output":true,"supports_system_messages":true,"supports_tool_choice":true,"source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-east-1/index.json","supports_audio_input":false,"supports_response_schema":true,"supports_vision":false,"supported_endpoints":["/v1/chat/completions"],"provider":"bedrock_mantle","base_model":"zai.glm-5"},"azure/us-gov/gpt-4.1":{"max_input_tokens":128000,"max_output_tokens":32768,"mode":"chat","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-4.1"},"azure/us-gov/gpt-4.1-mini":{"max_input_tokens":128000,"max_output_tokens":32768,"mode":"chat","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-4.1-mini"},"azure/us-gov/gpt-4o":{"max_input_tokens":128000,"max_output_tokens":16384,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-4o"},"azure/us-gov/gpt-5.6-luna":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-5.6-luna"},"azure/us-gov/gpt-5.6-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-5.6-sol"},"azure/us-gov/gpt-5.6-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_none_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/gpt-5.6-terra"},"azure/us-gov/text-embedding-ada-002":{"max_input_tokens":8192,"mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=armRegionName%20eq%20'usgovvirginia'%20and%20contains(productName,'OpenAI')","provider":"azure","base_model":"us-gov/text-embedding-ada-002"},"cloudflare/@cf/openai/whisper":{"mode":"audio_transcription","rpm":720,"source":"https://developers.cloudflare.com/workers-ai/models/whisper/","supported_endpoints":["/v1/audio/transcriptions"],"provider":"cloudflare","base_model":""},"cloudflare/@cf/openai/whisper-large-v3-turbo":{"mode":"audio_transcription","rpm":720,"source":"https://developers.cloudflare.com/workers-ai/models/whisper-large-v3-turbo/","supported_endpoints":["/v1/audio/transcriptions"],"provider":"cloudflare","base_model":""},"gemini/lyria-realtime-exp":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","source":"https://ai.google.dev/gemini-api/docs/models/lyria-realtime-exp","supported_modalities":["text"],"supported_output_modalities":["audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_function_calling":false,"supports_prompt_caching":false,"supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_web_search":false,"provider":"gemini","base_model":"lyria-realtime"},"perplexity/openai/gpt-5-nano":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-5-nano"},"perplexity/openai/gpt-5.3-codex":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-5.3-codex"},"perplexity/google/gemini-4-argon":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gemini-4-argon"},"perplexity/google/gemini-4-hydrogen":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gemini-4-hydrogen"},"perplexity/google/gemini-4-carbon":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gemini-4-carbon"},"perplexity/perplexity/nemotron-3.5-super-120b-a12b":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"nemotron-3.5-super-120b-a12b"},"baseten/glm-latest":{"alias_of":"baseten/zai-org/GLM-5.3","base_model":"glm"},"baseten/zai-org/GLM-5.3-Fast":{"max_input_tokens":1048576,"max_output_tokens":262144,"mode":"chat","source":"https://inference.baseten.co/v1/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"glm-5.3-fast"},"vertex_ai/gemini-2.5-flash-native-audio":{"deprecation_date":"2026-12-13","mode":"realtime","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-2.5-flash-native-audio"},"vertex_ai/gemini-2.5-flash-preview-tts":{"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-2.5-flash-tts"},"vertex_ai/gemini-3.1-flash-tts-preview":{"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-3.1-flash-tts"},"vertex_ai/gemini-3.5-transcribe":{"mode":"audio_transcription","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-3.5-transcribe"},"vertex_ai/gemini-3.5-transcribe-live":{"mode":"audio_transcription","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-3.5-transcribe-live"},"gpt-5.5-cyber-preview":{"deprecation_date":"2027-02-01","deprecation_source":"https://developers.openai.com/api/docs/deprecations","mode":"chat","fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_reasoning":true,"provider":"openai","base_model":"gpt-5.5-cyber","web_search_billing_unit":"per_request"},"together_ai/Salesforce/Llama-Rank-V1":{"deprecation_date":"2026-02-06","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":8192,"mode":"rerank","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"salesforce/llama-rank-v1","is_deprecated":true},"together_ai/meta-llama/Meta-Llama-3.1-8B":{"max_input_tokens":16384,"max_output_tokens":16384,"mode":"completion","source":"https://api.together.xyz/v1/models","provider":"together_ai","base_model":"llama-3.1-8b"},"together_ai/together/Tev1-4B-experimental":{"max_input_tokens":32768,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"together/tev1-4b"},"together_ai/Qwen/Qwen2.5-VL-72B-Instruct":{"deprecation_date":"2026-01-05","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":32768,"max_output_tokens":32768,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"qwen2.5-vl-72b-instruct","is_deprecated":true},"together_ai/Qwen/Qwen3-Coder-Next-FP8":{"deprecation_date":"2026-05-14","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"qwen3-coder-next-fp8","is_deprecated":true},"together_ai/Qwen/Qwen3-VL-32B-Instruct":{"deprecation_date":"2026-02-25","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"qwen3-vl-32b-instruct","is_deprecated":true},"together_ai/Qwen/Qwen3-VL-8B-Instruct":{"deprecation_date":"2026-04-16","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"qwen3-vl-8b-instruct","is_deprecated":true},"together_ai/Qwen/Qwen3.5-397B-A17B":{"deprecation_date":"2026-06-29","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"qwen3.5-397b-a17b","is_deprecated":true},"together_ai/essentialai/rnj-1-instruct":{"deprecation_date":"2026-11-14","mode":"chat","source":"https://www.together.ai/models/rnj-1-instruct","provider":"together_ai","base_model":"essentialai/rnj-1-instruct","is_deprecated":true},"together_ai/google/gemma-4-31B-it":{"deprecation_date":"2026-09-15","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","successor_model":"together_ai/zai-org/GLM-5.3-Flash","provider":"together_ai","base_model":"gemma-4-31b-it","is_deprecated":true},"together_ai/moonshotai/Kimi-K2.6":{"deprecation_date":"2026-08-19","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"kimi-k2.6","is_deprecated":true},"together_ai/moonshotai/Kimi-K2.7-Code":{"deprecation_date":"2026-08-27","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"kimi-k2.7-code","is_deprecated":true},"together_ai/nvidia/nemotron-3-ultra-550b-a55b":{"deprecation_date":"2026-08-27","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":512288,"max_output_tokens":512288,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"nemotron-3-ultra-550b-a55b","is_deprecated":true},"together_ai/zai-org/GLM-5":{"deprecation_date":"2026-06-22","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":202752,"max_output_tokens":202752,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"glm-5","is_deprecated":true},"together_ai/zai-org/GLM-5.1":{"deprecation_date":"2026-07-10","deprecation_source":"https://docs.together.ai/docs/deprecations","max_input_tokens":202752,"max_output_tokens":202752,"mode":"chat","source":"https://api.together.ai/v1/models","provider":"together_ai","base_model":"glm-5.1","is_deprecated":true},"azure/eu/o3":{"deprecation_date":"2026-11-19","successor_model":"azure/eu/gpt-5.6-sol","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"o3","supports_prompt_caching":true},"azure/eu/o4-mini":{"deprecation_date":"2026-11-19","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/eu/gpt-5.6-terra","max_input_tokens":200000,"max_output_tokens":100000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"o4-mini","supports_prompt_caching":true},"azure/eu/text-embedding-3-large":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-3-large"},"azure/eu/text-embedding-3-small":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-3-small"},"azure/eu/text-embedding-ada-002":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-ada-002"},"gemini/gemini-3.8-live":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_vision":true,"supports_web_search":true,"gemini_audio_only_live":true,"supports_response_schema":false,"provider":"gemini","base_model":"gemini-3.8-live"},"gemini/gemini-3.8-live-extended-thinking":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_vision":true,"supports_web_search":true,"gemini_audio_only_live":true,"supports_reasoning":true,"supports_response_schema":false,"provider":"gemini","base_model":"gemini-3.8-live-extended-thinking"},"azure/us/text-embedding-3-large":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-3-large"},"azure/us/text-embedding-3-small":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-3-small"},"azure/us/text-embedding-ada-002":{"deprecation_date":"2028-02-09","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","mode":"embedding","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"text-embedding-ada-002"},"azure/gpt-realtime-2":{"deprecation_date":"2027-01-20","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","successor_model":"azure/gpt-realtime-2.1","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-2","supports_prompt_caching":true},"azure/gpt-live-1":{"deprecation_date":"2027-09-10","mode":"realtime","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"azure","base_model":"gpt-live-1"},"azure/gpt-live-transcribe":{"deprecation_date":"2028-02-01","max_input_tokens":32000,"max_output_tokens":4096,"mode":"audio_transcription","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/realtime","/v1/realtime/transcription_sessions"],"supported_modalities":["text","audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider_specific_entry":{"realtime_ga_only":1},"supports_native_streaming":true,"provider":"azure","base_model":"gpt-live-transcribe"},"azure/gpt-transcribe":{"deprecation_date":"2028-02-01","mode":"audio_transcription","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/audio/transcriptions","/v1/realtime/transcription_sessions","/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"provider_specific_entry":{"realtime_ga_only":1,"transcription_json_or_text_only":1,"transcription_usage_duration":1},"supports_native_streaming":true,"provider":"azure","base_model":"gpt-transcribe"},"azure/gpt-realtime-translate":{"deprecation_date":"2027-11-09","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_modalities":["audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"provider_specific_entry":{"realtime_ga_only":1},"supports_native_streaming":true,"provider":"azure","base_model":"gpt-realtime-translate"},"aiand/deepseek-ai/deepseek-v4-flash":{"deprecation_date":"2026-10-27","max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"deepseek-v4-flash","is_deprecated":true},"aiand/deepseek-ai/deepseek-v4-pro":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"deepseek-v4-pro"},"aiand/deepseek-ai/deepseek-v4.1-flash":{"deprecation_date":"2026-11-05","max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"deepseek-v4.1-flash","is_deprecated":true},"aiand/deepseek-ai/deepseek-v4.1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"deepseek-v4.1-flash"},"aiand/google/gemma-4-31b-it":{"max_input_tokens":262144,"max_output_tokens":32768,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"gemma-4-31b-it"},"aiand/meta/muse-spark-open":{"max_input_tokens":1048576,"max_output_tokens":943718,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"muse-spark-open"},"aiand/moonshotai/kimi-k2.7-code":{"deprecation_date":"2026-10-27","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"kimi-k2.7-code","is_deprecated":true},"aiand/moonshotai/kimi-k3":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"kimi-k3"},"aiand/motif-technologies/motif-3":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"motif-technologies/motif-3"},"aiand/openai/gpt-oss-120b":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"gpt-oss-120b"},"aiand/qwen/qwen3.6-27b":{"max_input_tokens":262144,"max_output_tokens":65536,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"qwen3.6-27b"},"aiand/qwen/qwen3.8-27b":{"max_input_tokens":262144,"max_output_tokens":32768,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"qwen3.8-27b"},"aiand/zai-org/glm-5.2":{"deprecation_date":"2026-10-27","max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"glm-5.2","is_deprecated":true},"aiand/zai-org/glm-5.3":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":false,"provider":"aiand","base_model":"glm-5.3"},"aiand/zai-org/glm-5.3-flash":{"max_input_tokens":1048550,"max_output_tokens":131072,"mode":"chat","source":"https://api.aiand.com/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"aiand","base_model":"glm-5.3-flash"},"bespoke/nimble-latest":{"max_input_tokens":8192,"mode":"evaluation","source":"https://github.com/bespokelabsai/nimble","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"bespoke","base_model":"nimble"},"bespoke/nimble":{"max_input_tokens":8192,"mode":"evaluation","source":"https://ollama.com/library/nimble","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model under the name Ollama serves it as; infrastructure costs are paid separately"},"provider":"bespoke","base_model":"nimble"},"bespoke/bespokelabs/Bespoke-Nimble-9B":{"max_input_tokens":8192,"mode":"evaluation","source":"https://github.com/bespokelabsai/nimble","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"bespoke","base_model":"bespokelabs/bespoke-nimble-9b"},"laya/english":{"mode":"evaluation","source":"https://github.com/NandhaKishorM/laya","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"laya","base_model":"english"},"laya/multilingual":{"mode":"evaluation","source":"https://github.com/NandhaKishorM/laya","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"laya","base_model":"multilingual"},"laya/typed-decisions":{"mode":"evaluation","source":"https://github.com/NandhaKishorM/laya","supported_endpoints":["/v1/systemone"],"metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"laya","base_model":"typed-decisions"},"strands_decider/strands-decider-2B-hobson-v19":{"mode":"evaluation","source":"https://huggingface.co/StrandsAgents/strands-decider-2B-hobson-v19","metadata":{"notes":"Self-hosted decision model; infrastructure costs are paid separately"},"provider":"strands_decider","base_model":"strands-decider-2b-hobson-v19"},"nace/drex-v1.0":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/reference/models","supports_vision":true,"provider":"nace","base_model":"drex"},"nace/drex-v1.5":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/reference/models","supports_vision":true,"provider":"nace","base_model":"drex"},"nace/drex-v2.0":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/reference/models","provider":"nace","base_model":"drex"},"nace/drex-latest":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/reference/models","supports_vision":true,"provider":"nace","base_model":"drex"},"nace/drex-max":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/guides/drex-max","supports_vision":true,"provider":"nace","base_model":"drex-max"},"nace/drex-max-v1.1":{"max_input_tokens":262144,"mode":"evaluation","source":"https://console.nace.ai/docs/drex/guides/drex-max","supports_vision":true,"provider":"nace","base_model":"drex-max"},"typesafe/jev-1.13.0":{"max_input_tokens":65536,"mode":"decisions","source":"https://docs.typesafe.ai/models","provider":"typesafe","base_model":"jev-1.13.0","supports_decisions":true,"max_tokens":65536,"rpm":1200,"tpm":15000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_video_input":false,"supports_pdf_input":false,"supports_function_calling":false,"supports_tool_choice":false},"typesafe/jev-1.14.0":{"max_input_tokens":65536,"mode":"decisions","source":"https://docs.typesafe.ai/models","supports_box_questions":true,"provider":"typesafe","base_model":"jev-1.14.0","supports_decisions":true},"typesafe/jev-latest":{"max_input_tokens":65536,"mode":"decisions","source":"https://docs.typesafe.ai/models","supports_box_questions":true,"provider":"typesafe","base_model":"jev-1.13.0","supports_decisions":true,"max_tokens":65536,"rpm":1200,"tpm":15000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_video_input":false,"supports_pdf_input":false,"supports_function_calling":false,"supports_tool_choice":false},"typesafe/jev-preview":{"max_input_tokens":65536,"mode":"decisions","source":"https://docs.typesafe.ai/models","supports_box_questions":true,"provider":"typesafe","base_model":"jev-1.13.0","supports_decisions":true,"max_tokens":65536,"rpm":1200,"tpm":15000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_video_input":false,"supports_pdf_input":false,"supports_function_calling":false,"supports_tool_choice":false},"eu_jev/eu-jeff-1.0.0":{"max_input_tokens":32768,"mode":"evaluation","source":"https://eu-jev.com/docs","supported_endpoints":["/v1/systemone"],"provider":"eu_jev","base_model":"eu-jeff-1.0.0"},"eu_jev/eu-jeff-latest":{"max_input_tokens":32768,"mode":"evaluation","source":"https://eu-jev.com/docs","supported_endpoints":["/v1/systemone"],"provider":"eu_jev","base_model":"eu-jeff"},"wandb/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1049000,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"wandb","base_model":"deepseek-v4.1-flash"},"wandb/google/gemma-4-26B-A4B-it":{"max_input_tokens":262000,"max_output_tokens":262144,"mode":"chat","source":"https://wandb.ai/site/pricing/tokens/","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wandb","base_model":"gemma-4-26b-a4b-it"},"bedrock/kimi-latest":{"alias_of":"global.moonshotai.kimi-k3","base_model":"kimi"},"baseten/kimi-latest":{"alias_of":"baseten/moonshotai/Kimi-K3","base_model":"kimi"},"baseten/glm-flash-latest":{"alias_of":"baseten/zai-org/GLM-5.3-Flash","base_model":"glm-flash"},"baseten/deepseek-flash-latest":{"alias_of":"baseten/deepseek-ai/DeepSeek-V4-Flash-0731","base_model":"deepseek-flash"},"baseten/deepseek-pro-latest":{"alias_of":"baseten/deepseek-ai/DeepSeek-V4-Pro-0813","base_model":"deepseek-pro"},"fireworks_ai/accounts/fireworks/models/deepseek-v4-pro":{"deprecation_date":"2026-08-27","max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"deepseek-v4-pro","is_deprecated":true},"fireworks_ai/accounts/fireworks/models/minimax-m2p7":{"deprecation_date":"2026-08-27","max_input_tokens":196608,"max_output_tokens":196608,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"fireworks_ai","base_model":"minimax-m2.7","is_deprecated":true},"fireworks_ai/accounts/fireworks/routers/rune":{"max_input_tokens":262144,"mode":"evaluation","source":"https://docs.fireworks.ai","provider":"fireworks_ai","base_model":"accounts/fireworks/routers/rune"},"fireworks_ai/accounts/fireworks/models/ember-1":{"max_input_tokens":1048576,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"ember-1","max_tokens":1048576},"vertex_ai/meta/llama-3.3-70b-instruct-maas":{"deprecation_date":"2026-10-21","max_input_tokens":128000,"max_output_tokens":8192,"mode":"chat","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_modalities":["text"],"supported_output_modalities":["text","code"],"supports_function_calling":true,"supports_tool_choice":true,"provider":"vertex_ai","base_model":"llama-3.3-70b-instruct","is_deprecated":true},"vertex_ai/virtual-try-on-001":{"deprecation_date":"2027-03-15","mode":"image_generation","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_modalities":["image"],"supported_output_modalities":["image"],"provider":"vertex_ai","base_model":"virtual-try-on-001"},"vertex_ai/gemini-2.5-flash-tts":{"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-2.5-flash-tts"},"vertex_ai/gemini-2.5-pro-tts":{"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","provider":"vertex_ai","base_model":"gemini-2.5-pro-tts"},"mistral.pixtral-large-2502-v1:0":{"max_input_tokens":128000,"max_output_tokens":131072,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_function_calling":true,"supports_tool_choice":false,"provider":"bedrock","base_model":"pixtral-large"},"azure/MAI-Cyber-1-Flash":{"max_input_tokens":256000,"max_output_tokens":64000,"mode":"chat","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"mai-cyber-1-flash"},"azure/mistral-ocr-2505":{"mode":"ocr","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/mistral-ai/","provider":"azure","base_model":"mistral-ocr"},"azure/mistral-ocr-2512":{"mode":"ocr","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/mistral-ai/","provider":"azure","base_model":"mistral-ocr"},"azure/claude-sonnet-latest":{"alias_of":"azure_ai/claude-sonnet-5-5","base_model":"claude-sonnet"},"bedrock/claude-sonnet-latest":{"alias_of":"global.anthropic.claude-sonnet-5-5","base_model":"claude-sonnet"},"perplexity/anthropic/claude-sonnet-5-5":{"mode":"responses","supports_adaptive_thinking":true,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"claude-sonnet-5-5"},"perplexity/anthropic/claude-fable-5-2":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"claude-fable-5-2"},"perplexity/anthropic/claude-fable-5-1":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"claude-fable-5-1"},"perplexity/anthropic/claude-opus-5-5":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"claude-opus-5-5"},"perplexity/openai/gpt-6.1-sol":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-6.1-sol"},"perplexity/openai/gpt-6-sol":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-6-sol"},"perplexity/openai/gpt-6-astra":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-6-astra"},"perplexity/openai/gpt-6-luna":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gpt-6-luna"},"perplexity/google/gemini-3.8-flash":{"mode":"responses","supports_sampling_params":false,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"gemini-3.8-flash"},"perplexity/xai/grok-4.7":{"mode":"responses","source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"grok-4.7"},"vertex_ai/claude-sonnet-latest":{"alias_of":"vertex_ai/claude-sonnet-5-5","base_model":"claude-sonnet"},"prism/deepseek-v4.1-flash":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","source":"https://prisminference.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"prism","base_model":"deepseek-v4.1-flash"},"prism/deepseek-v4-flash":{"max_input_tokens":1000000,"max_output_tokens":384000,"mode":"chat","source":"https://prisminference.com/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"prism","base_model":"deepseek-v4-flash"},"global.xai.grok-4.7":{"supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-xai-grok-4-7.html","supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"grok-4.7"},"bedrock/grok-latest":{"alias_of":"global.xai.grok-4.7","base_model":"grok"},"us.xai.grok-4.7":{"supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-xai-grok-4-7.html","supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"grok-4.7"},"us-gov.xai.grok-4.7":{"supports_regex_lookaround":false,"supported_endpoints":["/v1/chat/completions"],"supports_bedrock_runtime_chat_completions_tools_with_reasoning":true,"supports_bedrock_runtime_chat_completions_response_format":true,"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supports_function_calling":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"grok-4.7"},"baseten/deepseek-ai/DeepSeek-V4.1-Flash-Fast":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"deepseek-v4.1-flash-fast"},"bedrock/gpt-sol-latest":{"alias_of":"global.openai.gpt-6.1-sol","base_model":"gpt-sol"},"bedrock_mantle/gpt-sol-latest":{"alias_of":"bedrock_mantle/openai.gpt-6.1-sol","base_model":"gpt-sol"},"vertex_ai/gemini-3.8-flash-tts":{"max_input_tokens":8192,"max_output_tokens":16384,"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"vertex_ai","base_model":"gemini-3.8-flash-tts"},"vertex_ai/gemini-3.8-flash-lite-tts":{"max_input_tokens":8192,"max_output_tokens":16384,"mode":"audio_speech","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/audio/speech"],"provider":"vertex_ai","base_model":"gemini-3.8-flash-lite-tts"},"vertex_ai/xai/grok-4.7":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","base_model":"grok-4.7"},"vertex_ai/grok-latest":{"alias_of":"vertex_ai/xai/grok-4.7","base_model":"grok"},"azure/kimi-k2-thinking":{"deprecation_date":"2026-03-29","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/retired-models","successor_model":"azure_ai/kimi-k2.5","max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/kimi/","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":false,"supports_vision":false,"provider":"azure","base_model":"kimi-k2-thinking","is_deprecated":true},"global.zai.glm-5.3":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","base_model":"global.zai.glm-5.3","supports_regex_lookaround":false,"supports_schema_pattern":false},"bedrock/glm-latest":{"alias_of":"global.zai.glm-5.3","base_model":"glm"},"us.zai.glm-5.3":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_audio_input":false,"supports_function_calling":true,"supports_native_structured_output":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"bedrock","base_model":"us.zai.glm-5.3","supports_regex_lookaround":false,"supports_schema_pattern":false},"amazon.nova-2-5-sonic":{"max_input_tokens":256000,"max_output_tokens":64000,"mode":"realtime","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-5-sonic.md","supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"bedrock","base_model":"nova-2-5-sonic"},"gemini/gemini-nano-banana-2.1":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"image_generation","rpm":1000,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"tpm":4000000,"web_search_billing_unit":"per_query","provider":"gemini","base_model":"gemini-nano-banana-2.1"},"gemini/nano-banana-latest":{"alias_of":"gemini/gemini-nano-banana-2.1","base_model":"nano-banana"},"gemini-nano-banana-2.1":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"image_generation","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-nano-banana-2.1"},"vertex_ai/gemini-nano-banana-2.1":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"image_generation","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/images/generations","/v1/images/edits","/v1/batch"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-nano-banana-2.1"},"vertex_ai/nano-banana-latest":{"alias_of":"vertex_ai/gemini-nano-banana-2.1","base_model":"nano-banana"},"azure/grok-4.7":{"deprecation_date":"2027-10-05","max_input_tokens":500000,"max_output_tokens":500000,"mode":"chat","source":"https://docs.x.ai/developers/models/grok-4.7","supports_function_calling":true,"supports_pdf_input":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"provider":"azure","base_model":"grok-4.7"},"azure/grok-latest":{"alias_of":"azure_ai/grok-4.7","base_model":"grok"},"in.moonshotai.kimi-k3":{"token_count_framing":64,"token_count_tokenizer":"kimi_k3","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3.md","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"kimi-k3"},"bedrock_mantle/claude-haiku-latest":{"alias_of":"bedrock_mantle/anthropic.claude-haiku-5-5","base_model":"claude-haiku"},"azure/claude-haiku-latest":{"alias_of":"azure_ai/claude-haiku-5-5","base_model":"claude-haiku"},"bedrock/claude-haiku-latest":{"alias_of":"global.anthropic.claude-haiku-5-5","base_model":"claude-haiku"},"perplexity/anthropic/claude-haiku-5-5":{"mode":"responses","supports_adaptive_thinking":true,"supports_web_search":true,"supports_function_calling":true,"source":"https://docs.perplexity.ai/docs/agent-api/models","provider":"perplexity","base_model":"claude-haiku-5-5","supports_native_effort":true},"vertex_ai/claude-haiku-latest":{"alias_of":"vertex_ai/claude-haiku-5-5","base_model":"claude-haiku"},"microsoft_365_copilot/chat":{"mode":"chat","supports_mid_conversation_system":true,"provider":"microsoft_365_copilot","base_model":"chat"},"us.twelvelabs.pegasus-1-5-v1:0":{"max_input_tokens":261120,"max_output_tokens":98304,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","base_model":"pegasus-1-5","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","af-south-1","mx-central-1","sa-east-1"],"max_tokens":98304},"global.twelvelabs.pegasus-1-5-v1:0":{"max_input_tokens":261120,"max_output_tokens":98304,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","base_model":"pegasus-1-5","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","af-south-1","mx-central-1","sa-east-1"],"max_tokens":98304},"twelvelabs.pegasus-1-5-v1:0":{"max_input_tokens":261120,"max_output_tokens":98304,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_video_input":true,"provider":"bedrock","base_model":"pegasus-1-5","supports_vision":false,"supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","af-south-1","mx-central-1","sa-east-1"],"max_tokens":98304},"azure/Microsoft-Decision-1":{"max_input_tokens":32768,"mode":"decisions","source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","provider":"azure","base_model":"microsoft-decision-1","supports_decisions":true},"gpt-rosalind-discovery":{"mode":"chat","source":"https://developers.openai.com/api/docs/pricing","provider":"openai","base_model":"gpt-rosalind-discovery","web_search_billing_unit":"per_request"},"chatgpt/gpt-rosalind-discovery":{"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses"],"provider":"chatgpt","base_model":"gpt-rosalind-discovery"},"azure/Cohere-Embed-V5-Fast":{"deprecation_date":"2027-01-23","max_input_tokens":128000,"mode":"embedding","output_vector_size":2048,"source":"https://prices.azure.com/api/retail/prices?$filter=productName%20eq%20'Cohere%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/embeddings"],"supported_modalities":["text","image"],"supports_embedding_image_input":true,"provider":"azure","base_model":"embed-v5-fast"},"azure/Cohere-Embed-V5-Pro":{"deprecation_date":"2027-01-23","max_input_tokens":128000,"mode":"embedding","output_vector_size":2048,"source":"https://prices.azure.com/api/retail/prices?$filter=productName%20eq%20'Cohere%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/embeddings"],"supported_modalities":["text","image"],"supports_embedding_image_input":true,"provider":"azure","base_model":"embed-v5-pro"},"scaledown/extract":{"mode":"chat","supports_response_schema":false,"supports_system_messages":false,"supports_vision":false,"supports_function_calling":false,"provider":"scaledown","base_model":"extract"},"scaledown/summarize":{"mode":"chat","supports_system_messages":true,"supports_vision":false,"supports_function_calling":false,"provider":"scaledown","base_model":"summarize"},"scaledown/compress":{"mode":"chat","supports_system_messages":true,"supports_function_calling":false,"supports_vision":false,"provider":"scaledown","base_model":"compress"},"scaledown/classify":{"mode":"chat","supports_system_messages":false,"supports_vision":false,"supports_function_calling":false,"provider":"scaledown","base_model":"classify"},"scaledown/decisions":{"mode":"chat","supports_system_messages":false,"supports_vision":false,"supports_function_calling":false,"provider":"scaledown","base_model":"decisions"},"vertex_ai/gemini-robotics-er-2-preview":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"chat","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-robotics-er-2"},"vertex_ai/gemini-4-argon-cyber-permissive":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","prompt_cache_min_tokens":6144,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-4-argon-cyber-permissive"},"github_copilot/gemini-4-argon":{"deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":200000,"max_output_tokens":65536,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/gemini-4-carbon","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"github_copilot","base_model":"gemini-4-argon"},"openinfer/@oi/GLM-5.3-Flash":{"max_input_tokens":393216,"max_output_tokens":393216,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"source":"https://api.openinfer.ai/v1/models","provider":"openinfer","base_model":""},"openinfer/@oi/Qwen3.5-27B":{"max_input_tokens":32768,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"source":"https://api.openinfer.ai/v1/models","provider":"openinfer","base_model":""},"openinfer/@oi/Qwen3.5-9B":{"max_input_tokens":100000,"max_output_tokens":100000,"mode":"chat","supports_function_calling":true,"supports_reasoning":true,"source":"https://api.openinfer.ai/v1/models","provider":"openinfer","base_model":""},"xai/grok-voice-3.0":{"mode":"realtime","source":"https://docs.x.ai/developers/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"xai","base_model":"grok-voice-3.0"},"xai/grok-voice-latest":{"mode":"realtime","source":"https://docs.x.ai/developers/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"xai","base_model":"grok-voice"},"xai/grok-voice-2.0":{"mode":"realtime","source":"https://docs.x.ai/developers/pricing","successor_model":"xai/grok-voice-3.0","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"provider":"xai","base_model":"grok-voice-2.0"},"together_ai/nvidia/NVIDIA-Nemotron-3.5-Super-VL":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","base_model":"nvidia-nemotron-3.5-super-vl"},"vertex_ai/deepseek-ai/deepseek-v4.1-flash-maas":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_regions":["global"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"vertex_ai","base_model":"deepseek-v4.1-flash"},"azure/claude-opus-5-5-fast":{"deprecation_date":"2027-01-15","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://learn.microsoft.com/en-us/azure/foundry/concepts/model-lifecycle-retirement","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"thinking_always_on":true,"provider":"azure","base_model":"claude-opus-5-5-fast"},"baseten/meta-models/muse-spark-open":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"models/muse-spark-open"},"baseten/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"nvidia-nemotron-3.5-super-vl-120b-a12b"},"wandb/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B":{"max_input_tokens":1049000,"max_output_tokens":943718,"mode":"chat","source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wandb","base_model":"nvidia-nemotron-3.5-super-vl-120b-a12b"},"together_ai/reflection/Beam-1.0-501B-A23B":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_response_schema":true,"provider":"together_ai","base_model":"reflection/beam-1.0-501b-a23b"},"baseten/reflection/Beam-1.0":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"provider":"baseten","base_model":"reflection/beam-1.0"},"nebius/reflection/beam":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://tokenfactory.nebius.com/api/public/models_info","supports_function_calling":true,"supports_reasoning":true,"provider":"nebius","base_model":"reflection/beam"},"baseten/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"max_input_tokens":1048576,"max_output_tokens":262144,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"deepseek-v4.1-flash"},"baseten/deepseek-ai/DeepSeek-V4.1-Flash-1017-Fast":{"max_input_tokens":1048576,"max_output_tokens":262144,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"deepseek-v4.1-flash-1017-fast"},"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"deepseek-v4.1-flash","max_tokens":1048576},"fireworks_ai/accounts/fireworks/routers/deepseek-flash-latest":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/deepseek-flash"},"github_copilot/claude-sonnet-4.6":{"deprecation_date":"2026-09-01","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_legacy_thinking":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"claude-sonnet-4-6","is_deprecated":true},"github_copilot/gpt-5.4-nano":{"max_input_tokens":272000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-5.4-nano"},"github_copilot/grok-4.5":{"deprecation_date":"2026-10-19","max_input_tokens":500000,"max_output_tokens":500000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"grok-4.5","is_deprecated":true},"github_copilot/grok-4.6":{"deprecation_date":"2026-12-14","deprecation_source":"https://github.blog/changelog/2026-11-13-upcoming-deprecation-of-selected-github-copilot-models-in-december","max_input_tokens":500000,"max_output_tokens":500000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","successor_model":"github_copilot/grok-4.7","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"grok-4.6"},"github_copilot/grok-4.7":{"max_input_tokens":500000,"max_output_tokens":500000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"grok-4.7"},"github_copilot/grok-latest":{"alias_of":"github_copilot/grok-4.7","base_model":"grok"},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","base_model":"deepseek-v4.1-flash"},"meta/muse-image-1.0":{"mode":"image_generation","source":"https://dev.meta.ai/docs/pricing-rate-limits","supported_endpoints":["/v1/images/generations","/v1/images/edits","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"meta","base_model":"muse-image-1.0"},"fireworks_ai/accounts/fireworks/models/minimax-m3p1-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"minimax-m3.1-flash"},"meta/sam-3.1":{"mode":"responses","source":"https://dev.meta.ai/docs/pricing-rate-limits.md","supported_endpoints":["/v1/responses"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"provider":"meta","base_model":"sam-3.1"},"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash-1017-us":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://docs.fireworks.ai/serverless/us-only-serverless","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"deepseek-v4.1-flash-1017-us"},"wandb/reflection/Beam-1.0-501B-A23B":{"max_input_tokens":1049000,"mode":"chat","source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wandb","base_model":"reflection/beam-1.0-501b-a23b"},"wandb/zai-org/GLM-5.3":{"max_input_tokens":1049000,"mode":"chat","source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wandb","base_model":"glm-5.3"},"azure/MAI-Code-1.1-Flash":{"deprecation_date":"2026-12-09","max_input_tokens":256000,"max_output_tokens":64000,"mode":"chat","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"mai-code-1.1-flash"},"opencode-go/grok-4.7":{"mode":"responses","supported_endpoints":["/v1/responses"],"max_output_tokens":500000,"max_input_tokens":500000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"tiered_pricing":[{"range":[0,200000],"input_cost_per_token":0.000002,"output_cost_per_token":0.000006,"cache_read_input_token_cost":5e-7},{"range":[200000,500000],"input_cost_per_token":0.000004,"output_cost_per_token":0.000012,"cache_read_input_token_cost":0.000001}],"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"grok-4.7"},"opencode-go/gpt-6-luna":{"mode":"responses","supported_endpoints":["/v1/responses"],"max_output_tokens":128000,"max_input_tokens":922000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"tiered_pricing":[{"range":[0,272000],"input_cost_per_token":1.0000000000000001e-7,"output_cost_per_token":5e-7,"cache_read_input_token_cost":1e-8,"cache_creation_input_token_cost":1.25e-7},{"range":[272000,1050000],"input_cost_per_token":2.0000000000000002e-7,"output_cost_per_token":7.5e-7,"cache_read_input_token_cost":2e-8,"cache_creation_input_token_cost":2.5e-7}],"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"gpt-6-luna"},"opencode-go/gpt-5.6-luna":{"mode":"responses","supported_endpoints":["/v1/responses"],"max_output_tokens":128000,"max_input_tokens":922000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"tiered_pricing":[{"range":[0,272000],"input_cost_per_token":2.0000000000000002e-7,"output_cost_per_token":0.0000012,"cache_read_input_token_cost":2e-8,"cache_creation_input_token_cost":2.5e-7},{"range":[272000,1050000],"input_cost_per_token":4.0000000000000003e-7,"output_cost_per_token":0.0000018000000000000001,"cache_read_input_token_cost":4e-8,"cache_creation_input_token_cost":5e-7}],"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"gpt-5.6-luna"},"opencode-go/claude-haiku-5-5":{"mode":"chat","supported_endpoints":["/v1/messages"],"max_output_tokens":128000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"tiered_pricing":[{"range":[0,100000],"input_cost_per_token":1.0000000000000001e-7,"output_cost_per_token":5e-7,"cache_read_input_token_cost":1e-8,"cache_creation_input_token_cost":1.25e-7},{"range":[100000,1000000],"input_cost_per_token":5e-7,"output_cost_per_token":0.0000025,"cache_read_input_token_cost":5.0000000000000004e-8,"cache_creation_input_token_cost":6.25e-7}],"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"claude-haiku-5-5"},"opencode-go/gemini-4-argon":{"mode":"chat","supported_endpoints":["/v1beta/models"],"max_output_tokens":65536,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"gemini-4-argon"},"opencode-go/glm-5.3-flash":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"glm-5.3-flash"},"opencode-go/glm-5.3":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"glm-5.3"},"opencode-go/kimi-k3.1-pro":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"kimi-k3.1-pro"},"opencode-go/kimi-k3":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"kimi-k3"},"opencode-go/longcat-2.0":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"longcat-2.0"},"opencode-go/deepseek-v4.1-flash":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":384000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"deepseek-v4.1-flash"},"opencode-go/mimo-v2.6-flash":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"mimo-v2.6-flash"},"opencode-go/mimo-v2.6-pro":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"mimo-v2.6-pro"},"opencode-go/minimax-m3.1-flash":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":512000,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"minimax-m3.1-flash"},"opencode-go/muse-spark-1.3-contributor":{"mode":"responses","supported_endpoints":["/v1/responses"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"muse-spark-1.3-contributor"},"opencode-go/muse-spark-1.2-contributor":{"mode":"responses","supported_endpoints":["/v1/responses"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"muse-spark-1.2-contributor"},"opencode-go/qwen3.8-max":{"mode":"chat","supported_endpoints":["/v1/messages"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"qwen3.8-max"},"opencode-go/qwen3.8-flash":{"mode":"chat","supported_endpoints":["/v1/messages"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"qwen3.8-flash"},"opencode-go/hy4-preview":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":64000,"max_input_tokens":1024000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"hy4"},"opencode-go/hy3":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":128000,"max_input_tokens":192000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"hy3"},"opencode-go/longcat-2.5-preview-free":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"longcat-2.5-free"},"opencode-go/step-5-preview":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":65536,"max_input_tokens":1000000,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"step-5"},"opencode-go/beam-free":{"mode":"chat","supported_endpoints":["/v1/chat/completions"],"max_output_tokens":131072,"max_input_tokens":1048576,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":false,"source":"https://opencode.ai/v2/docs/console/go/","provider":"opencode-go","base_model":"beam-free"},"bedrock/claude_platform/claude-haiku-4-5":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_computer_use":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":4096,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-haiku-4-5"},"bedrock/claude_platform/claude-opus-4-5":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":4096,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-opus-4-5"},"bedrock/claude_platform/claude-opus-4-6":{"supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"supports_max_reasoning_effort":true,"prompt_cache_min_tokens":4096,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-opus-4-6"},"bedrock/claude_platform/claude-opus-4-7":{"supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":2048,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-opus-4-7"},"bedrock/claude_platform/claude-opus-4-8":{"supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-opus-4-8"},"bedrock/claude_platform/claude-sonnet-4-5":{"deprecation_date":"2026-11-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","successor_model":"bedrock/claude_platform/claude-sonnet-5-5","max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_tool_choice":true,"supports_vision":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-sonnet-4-5"},"bedrock/claude_platform/claude-sonnet-4-6":{"supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"supports_legacy_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_max_reasoning_effort":true,"supports_tool_choice":true,"supports_vision":true,"supports_output_config":true,"prompt_cache_min_tokens":1024,"provider_specific_entry":{"us":1.1},"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-sonnet-4-6"},"bedrock/claude_platform/claude-sonnet-5":{"supports_web_search":true,"supports_anthropic_compaction":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_thinking_cache_preservation":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"provider_specific_entry":{"us":1.1},"supports_output_config":true,"prompt_cache_min_tokens":1024,"source":"https://platform.claude.com/docs/en/about-claude/pricing#claude-platform-on-aws-pricing","provider":"bedrock","base_model":"claude-platform/claude-sonnet-5"},"chatgpt/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6.1-terra"},"bedrock/gpt-terra-latest":{"alias_of":"global.openai.gpt-6.1-terra","base_model":"gpt-terra"},"baseten/decision-models/baseten-elephant-9b-latest":{"max_input_tokens":16384,"mode":"decisions","source":"https://inference.baseten.co/v1/models","supports_vision":true,"provider":"baseten","base_model":"decision-models/baseten-elephant-9b"},"baseten/decision-models/baseten-mammoth-27b-latest":{"max_input_tokens":16384,"mode":"decisions","source":"https://inference.baseten.co/v1/models","supports_vision":true,"provider":"baseten","base_model":"decision-models/baseten-mammoth-27b"},"github_copilot/gpt-6.1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://github.blog/changelog/2026-10-26-gpt-6-1-terra-in-github-copilot","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6.1-terra"},"github_copilot/gpt-terra-latest":{"alias_of":"github_copilot/gpt-6.1-terra","base_model":"gpt-terra"},"refuseless/glm-5.3-abliterated":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://refuseless.com/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["low","high","max"],"provider":"refuseless","base_model":"glm-5.3-abliterated"},"refuseless/glm-5.3-flash-abliterated":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","source":"https://refuseless.com/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["low","high","max"],"provider":"refuseless","base_model":"glm-5.3-flash-abliterated"},"refuseless/deepseek-v4.1-flash-abliterated":{"max_input_tokens":1048576,"max_output_tokens":130000,"mode":"chat","source":"https://refuseless.com/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["low","high","max"],"provider":"refuseless","base_model":"deepseek-v4.1-flash-abliterated"},"refuseless/deepseek-v4.1-flash-fast":{"max_input_tokens":1048576,"max_output_tokens":262144,"mode":"chat","source":"https://refuseless.com/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["low","high","max"],"provider":"refuseless","base_model":"deepseek-v4.1-flash-fast"},"refuseless/minimax-m3.1-flash":{"max_input_tokens":1048576,"max_output_tokens":512000,"mode":"chat","source":"https://refuseless.com/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"reasoning_effort_levels":["low","high","max"],"provider":"refuseless","base_model":"minimax-m3.1-flash"},"fireworks_ai/accounts/fireworks/models/rune":{"max_input_tokens":262144,"mode":"decisions","source":"https://api.fireworks.ai/v1/serverless/models?format=nested","provider":"fireworks_ai","base_model":"rune"},"bedrock_mantle/gpt-terra-latest":{"alias_of":"bedrock_mantle/openai.gpt-6.1-terra","base_model":"gpt-terra"},"bedrock_mantle/openai.gpt-6.2-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/responses","/v1/chat/completions"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"gpt-6.2-sol","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_sampling_params":false,"supports_service_tier":false,"supports_system_messages":true,"supports_tool_search":true,"supports_web_search":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock_mantle/openai.gpt-6.2-astra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/responses","/v1/chat/completions"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"gpt-6.2-astra","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default","ultrafast"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_sampling_params":false,"supports_service_tier":true,"supports_system_messages":true,"supports_tool_search":true,"supports_web_search":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"clf_ai_gateway/glm-5.3":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"glm-5.3"},"clf_ai_gateway/glm-5.3-flash":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_vision":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"glm-5.3-flash"},"clf_ai_gateway/glm-5.2":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":262144,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"glm-5.2"},"clf_ai_gateway/glm-4.7-flash":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"supports_response_schema":true,"supports_system_messages":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"glm-4.7-flash"},"clf_ai_gateway/deepseek-v4-pro":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"deepseek-v4-pro"},"clf_ai_gateway/deepseek-v4-flash":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"deepseek-v4-flash"},"clf_ai_gateway/deepseek-v4.1-flash":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_vision":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"deepseek-v4.1-flash"},"clf_ai_gateway/kimi-k2.6":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":262144,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","medium","high"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_vision":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"kimi-k2.6"},"clf_ai_gateway/kimi-k2.7-code":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":262144,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_vision":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"kimi-k2.7-code"},"clf_ai_gateway/qwen3.8-27b":{"metadata":{"comment":"Billed price checked 2026-10-29, which the gateway shows as 60% off a list price 2.5x higher. Update the rows when the discount ends."},"max_input_tokens":262144,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","xhigh"],"supports_response_schema":true,"supports_system_messages":true,"supports_prompt_caching":true,"supports_vision":true,"source":"https://clfaigateway.dev/models","provider":"clf_ai_gateway","base_model":"qwen3.8-27b"},"kyma/claude-fable-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"provider":"kyma","base_model":"claude-fable-5"},"kyma/claude-fable-5.1":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"provider":"kyma","base_model":"claude-fable-5-1"},"kyma/claude-haiku-4-5":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"prompt_cache_min_tokens":4096,"provider":"kyma","base_model":"claude-haiku-4-5"},"kyma/claude-haiku-5.5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"provider":"kyma","base_model":"claude-haiku-5-5"},"kyma/claude-opus-4-7":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"supports_adaptive_thinking":true,"prompt_cache_min_tokens":2048,"provider":"kyma","base_model":"claude-opus-4-7"},"kyma/claude-opus-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"prompt_cache_min_tokens":512,"provider":"kyma","base_model":"claude-opus-5"},"kyma/claude-opus-5.5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"prompt_cache_min_tokens":512,"provider":"kyma","base_model":"claude-opus-5-5"},"kyma/claude-sonnet-4-6":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"supports_adaptive_thinking":true,"prompt_cache_min_tokens":1024,"provider":"kyma","base_model":"claude-sonnet-4-6"},"kyma/claude-sonnet-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"prompt_cache_min_tokens":1024,"provider":"kyma","base_model":"claude-sonnet-5"},"kyma/claude-sonnet-5.5":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"prompt_cache_min_tokens":512,"provider":"kyma","base_model":"claude-sonnet-5-5"},"kyma/deepseek-r1":{"max_input_tokens":64000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"deepseek-r1"},"kyma/deepseek-v3":{"max_input_tokens":160000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"deepseek-v3"},"kyma/deepseek-v4-flash":{"max_input_tokens":1000000,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"deepseek-v4-flash"},"kyma/deepseek-v4-flash-vision-exp":{"max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"deepseek-v4-flash-vision"},"kyma/deepseek-v4.1-flash":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"deepseek-v4.1-flash"},"kyma/gemini-3-flash":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gemini-3-flash"},"kyma/gemini-3.1-pro":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gemini-3.1-pro"},"kyma/gemini-3.5-flash":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gemini-3.5-flash"},"kyma/gemini-3.5-flash-lite":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gemini-3.5-flash-lite"},"kyma/gemini-3.6-flash":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_pdf_input":false,"provider":"kyma","base_model":"gemini-3.6-flash"},"kyma/gemini-3.7-flash":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gemini-3.7-flash"},"kyma/gemini-3.8-flash":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gemini-3.8-flash"},"kyma/gemini-4-argon":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gemini-4-argon"},"kyma/gemma-4-31b":{"max_input_tokens":128000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gemma-4-31b"},"kyma/glm-4.5-air":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-4.5-air"},"kyma/glm-4.7-flash":{"max_input_tokens":203000,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-4.7-flash"},"kyma/glm-5.1":{"max_input_tokens":203000,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-5.1"},"kyma/glm-5.2":{"max_input_tokens":1000000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-5.2"},"kyma/glm-5.3":{"max_input_tokens":1310720,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-5.3"},"kyma/glm-5.3-flash":{"max_input_tokens":1310720,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"glm-5.3-flash"},"kyma/gpt-5.6-luna":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-5.6-luna"},"kyma/gpt-5.6-luna-pro":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-5.6-luna-pro"},"kyma/gpt-5.6-sol":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-5.6-sol"},"kyma/gpt-5.6-sol-pro":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-5.6-sol-pro"},"kyma/gpt-5.6-terra":{"max_input_tokens":1050000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gpt-5.6-terra"},"kyma/gpt-5.6-terra-pro":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-5.6-terra-pro"},"kyma/gpt-6-astra":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gpt-6-astra"},"kyma/gpt-6-luna":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-6-luna"},"kyma/gpt-6.1-sol":{"max_input_tokens":1050000,"max_output_tokens":128000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"gpt-6.1-sol"},"kyma/gpt-oss-120b":{"max_input_tokens":128000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"gpt-oss-120b"},"kyma/grok-4.20":{"max_input_tokens":2000000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"grok-4.20"},"kyma/grok-4.20-multi-agent":{"max_input_tokens":2000000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":false,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"grok-4.20-multi-agent"},"kyma/grok-4.3":{"max_input_tokens":1000000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"grok-4.3"},"kyma/grok-4.5":{"max_input_tokens":500000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"grok-4.5"},"kyma/grok-4.6":{"max_input_tokens":500000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"grok-4.6"},"kyma/grok-4.7":{"max_input_tokens":500000,"max_output_tokens":450000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"grok-4.7"},"kyma/grok-build":{"max_input_tokens":256000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"grok-build"},"kyma/hy3":{"max_input_tokens":262144,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"hy3"},"kyma/hy4-preview":{"max_input_tokens":1048576,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"hy4"},"kyma/kimi-k2.6":{"max_input_tokens":262144,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"kimi-k2.6"},"kyma/kimi-k2.7-code":{"max_input_tokens":262144,"max_output_tokens":16384,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"kimi-k2.7-code"},"kyma/kimi-k3":{"max_input_tokens":1048576,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"kimi-k3"},"kyma/kimi-k3.1-pro":{"max_input_tokens":1048576,"max_output_tokens":943718,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"kimi-k3.1-pro"},"kyma/llama-3.3-70b":{"max_input_tokens":128000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"llama-3.3-70b"},"kyma/llama-4-maverick":{"max_input_tokens":1048576,"max_output_tokens":16384,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"llama-4-maverick"},"kyma/mimo-v2.5":{"max_input_tokens":1050000,"max_output_tokens":16384,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"mimo"},"kyma/mimo-v2.6-flash":{"max_input_tokens":1050000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"mimo-v2.6-flash"},"kyma/mimo-v2.6-pro":{"max_input_tokens":1050000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"mimo-v2.6-pro"},"kyma/minimax-m2.5":{"max_input_tokens":196608,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"minimax-m2.5"},"kyma/minimax-m2.7":{"max_input_tokens":204800,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"minimax-m2.7"},"kyma/minimax-m3":{"max_input_tokens":1048576,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"minimax-m3"},"kyma/minimax-m3.1-flash":{"max_input_tokens":1048576,"max_output_tokens":524288,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"minimax-m3.1-flash"},"kyma/muse-glimmer-30b":{"max_input_tokens":131072,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"muse-glimmer-30b"},"kyma/muse-spark-1.1":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"muse-spark-1.1"},"kyma/muse-spark-1.2":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"muse-spark-1.2"},"kyma/muse-spark-1.3":{"max_input_tokens":1048576,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"muse-spark-1.3"},"kyma/muse-spark-open":{"max_input_tokens":1048576,"max_output_tokens":943718,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"muse-spark-open"},"kyma/muse-wonder":{"max_input_tokens":1048576,"max_output_tokens":943718,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":true,"provider":"kyma","base_model":"muse-wonder"},"kyma/nemotron-3-ultra-550b":{"max_input_tokens":1000000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"nemotron-3-ultra-550b"},"kyma/nemotron-3.5-super-vl":{"max_input_tokens":1048576,"max_output_tokens":235929,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"nemotron-3.5-super-vl"},"kyma/qwen-3-32b":{"max_input_tokens":32768,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3-32b"},"kyma/qwen-3-coder":{"max_input_tokens":131072,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3-coder"},"kyma/qwen-3.6-plus":{"max_input_tokens":1000000,"max_output_tokens":65536,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.6-plus"},"kyma/qwen-3.7-max":{"max_input_tokens":1000000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.7-max"},"kyma/qwen-3.7-plus":{"max_input_tokens":1000000,"max_output_tokens":32768,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.7-plus"},"kyma/qwen-3.8-max":{"max_input_tokens":1000000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.8-max"},"kyma/qwen3.7-flash":{"max_input_tokens":1000000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.7-flash"},"kyma/qwen3.8-27b":{"max_input_tokens":1000000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.8-27b"},"kyma/qwen3.8-flash":{"max_input_tokens":1000000,"max_output_tokens":131072,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"qwen3.8-flash"},"kyma/sonar":{"max_input_tokens":127072,"max_output_tokens":4096,"mode":"chat","supports_function_calling":false,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_pdf_input":false,"provider":"kyma","base_model":"sonar"},"kyma/sonar-pro":{"max_input_tokens":200000,"max_output_tokens":8000,"mode":"chat","supports_function_calling":false,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":false,"supports_pdf_input":false,"provider":"kyma","base_model":"sonar-pro"},"kyma/step-3.7-flash":{"max_input_tokens":256000,"max_output_tokens":8192,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"step-3.7-flash"},"kyma/step-5-preview":{"max_input_tokens":1048576,"max_output_tokens":64000,"mode":"chat","supports_function_calling":true,"supports_vision":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_pdf_input":false,"provider":"kyma","base_model":"step-5"},"databricks/databricks-claude-haiku-5-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Endpoint databricks-claude-haiku-5-5. Capabilities from the Databricks supported-models page; context window and max output follow Anthropic's published limits. Pricing is Anthropic's first-party list price (https://platform.claude.com/docs/en/about-claude/pricing), tiered at 100k prompt tokens. Databricks had not published DBU rates for Haiku 5.5 as of 2026-10-08, so Databricks billing may differ."},"mode":"chat","prompt_cache_min_tokens":512,"source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models#-anthropic-claude-haiku-55","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_anthropic_thinking_payload":true,"supports_assistant_prefill":false,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":false,"provider":"databricks","base_model":"claude-haiku-5-5","supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","tool_use_system_prompt_tokens":286,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_image_input":true,"supports_system_messages":true,"supports_reasoning_disable":true,"supports_reasoning_effort":false,"reasoning_effort_levels":["low","medium","high"],"default_reasoning_effort":"medium","supports_xhigh_reasoning_effort":false,"supports_max_reasoning_effort":false},"databricks/claude-haiku-latest":{"alias_of":"databricks/databricks-claude-haiku-5-5","base_model":"claude-haiku"},"databricks/databricks-claude-sonnet-5-5":{"max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_anthropic_thinking_payload":true,"supports_assistant_prefill":false,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":false,"thinking_always_on":true,"provider":"databricks","base_model":"claude-sonnet-5-5"},"databricks/claude-sonnet-latest":{"alias_of":"databricks/databricks-claude-sonnet-5-5","base_model":"claude-sonnet"},"databricks/databricks-deepseek-v4-1-flash":{"deprecation_date":"2026-11-19","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","max_input_tokens":1000000,"max_output_tokens":384000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields. The docs list only databricks-deepseek-v4-1-flash (retiring 2026-11-19) and databricks-deepseek-v4-1-flash-1017, so the pricing page's 'DeepSeek V4.1 Flash (0910)' row prices the unsuffixed endpoint."},"mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"successor_model":"databricks/databricks-deepseek-v4-1-flash-1017","provider":"databricks","base_model":"deepseek-v4-1-flash"},"databricks/databricks-deepseek-v4-1-flash-1017":{"max_input_tokens":1000000,"max_output_tokens":384000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"deepseek-v4-1-flash"},"databricks/databricks-gemini-2-5-flash-lite":{"max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"databricks","base_model":"gemini-2.5-flash-lite"},"databricks/databricks-gemini-3-1-flash-lite-image":{"max_input_tokens":65536,"max_output_tokens":4096,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text","image"],"supports_function_calling":false,"supports_vision":true,"provider":"databricks","base_model":"gemini-3.1-flash-lite-image"},"databricks/databricks-gemini-4-argon":{"max_input_tokens":1048576,"max_output_tokens":1048576,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"databricks","base_model":"gemini-4-argon"},"databricks/databricks-gpt-5-4-pro":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"responses","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-5-4-pro"},"databricks/databricks-gpt-6-1-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"responses","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-6-1-sol"},"databricks/gpt-sol-latest":{"alias_of":"databricks/databricks-gpt-6-1-sol","base_model":"gpt-sol"},"databricks/databricks-gpt-6-1-terra":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-6-1-terra"},"databricks/gpt-terra-latest":{"alias_of":"databricks/databricks-gpt-6-1-terra","base_model":"gpt-terra"},"databricks/databricks-gpt-6-astra":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-6-astra"},"databricks/gpt-astra-latest":{"alias_of":"databricks/databricks-gpt-6-astra","base_model":"gpt-astra"},"databricks/databricks-gpt-6-luna":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"responses","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-6-luna"},"databricks/gpt-luna-latest":{"alias_of":"databricks/databricks-gpt-6-luna","base_model":"gpt-luna"},"databricks/databricks-gpt-6-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"responses","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-6-sol"},"databricks/databricks-gpt-image-1":{"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","provider":"databricks","base_model":"gpt-image-1"},"databricks/databricks-gpt-image-1-5":{"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_pdf_input":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-image-1-5"},"databricks/databricks-gpt-image-1-mini":{"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","provider":"databricks","base_model":"gpt-image-1-mini"},"databricks/databricks-gpt-image-2":{"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_pdf_input":true,"supports_vision":true,"provider":"databricks","base_model":"gpt-image-2"},"databricks/databricks-grok-4-7":{"max_input_tokens":500000,"max_output_tokens":500000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"grok-4-7"},"databricks/grok-latest":{"alias_of":"databricks/databricks-grok-4-7","base_model":"grok"},"databricks/databricks-kimi-k2-7":{"deprecation_date":"2026-10-30","deprecation_source":"https://docs.databricks.com/aws/en/machine-learning/retired-models-policy","max_input_tokens":262144,"max_output_tokens":262144,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","successor_model":"databricks/databricks-kimi-k3","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"provider":"databricks","base_model":"kimi-k2-7","is_deprecated":true},"fireworks_ai/accounts/fireworks/routers/spot-deepseek-v4-flash":{"deprecation_date":"2026-08-14","deprecation_source":"https://docs.fireworks.ai/api-reference/list-models.md","max_input_tokens":1048576,"max_output_tokens":384000,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","successor_model":"fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/spot-deepseek-v4-flash","is_deprecated":true},"fireworks_ai/accounts/fireworks/routers/kimi-k3-spot":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/kimi-k3-spot"},"fireworks_ai/accounts/fireworks/routers/glm-5p3-spot":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/glm-5.3-spot"},"fireworks_ai/accounts/fireworks/routers/deepseek-v4p1-flash-spot":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/deepseek-v4.1-flash-spot"},"fireworks_ai/accounts/fireworks/routers/deepseek-v4p1-flash-1017-ultrafast":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing.md","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/deepseek-v4.1-flash-1017-ultrafast"},"fireworks_ai/accounts/fireworks/routers/deepseek-v4p1-flash-1017-spot":{"max_input_tokens":1048576,"max_output_tokens":393216,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"accounts/fireworks/routers/deepseek-v4.1-flash-1017-spot"},"bigmodel/glm-5.3":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.bigmodel.cn/cn/guide/start/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"thinking_always_on":true,"metadata":{"comment":"Official API list price checked 2026-10-27: input 8, output 28, cached input 2 CNY per million tokens; converted at 7.0 CNY/USD. Cache storage is temporarily free and is not encoded."},"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_native_streaming":true,"provider":"bigmodel","base_model":"glm-5.3"},"bigmodel/glm-5.3-flash":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.bigmodel.cn/cn/guide/start/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"thinking_always_on":true,"supports_vision":true,"supports_video_input":true,"metadata":{"comment":"Official API list price checked 2026-10-27: input 0.8, output 2.8, cached input 0.23 CNY per million tokens; converted at 7.0 CNY/USD. Cache storage is temporarily free and is not encoded."},"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_native_streaming":true,"provider":"bigmodel","base_model":"glm-5.3-flash"},"bigmodel/glm-5.3-flashx":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","source":"https://docs.bigmodel.cn/cn/guide/start/pricing","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"thinking_always_on":true,"supports_vision":true,"supports_video_input":true,"metadata":{"comment":"Official API list price checked 2026-10-27: input 2, output 7, cached input 0.57 CNY per million tokens; converted at 7.0 CNY/USD. Cache storage is temporarily free and is not encoded."},"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages"],"supports_native_streaming":true,"provider":"bigmodel","base_model":"glm-5.3-flashx"},"fireworks_ai/accounts/fireworks/models/ultramarine-1-flash-us":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://docs.fireworks.ai/serverless/pricing.md","supports_function_calling":false,"supports_prompt_caching":true,"supports_response_schema":true,"supports_tool_choice":false,"supports_vision":true,"provider":"fireworks_ai","base_model":"ultramarine-1-flash-us"},"baseten/decision-models/baseten-joey-2b-latest":{"max_input_tokens":16384,"mode":"decisions","source":"https://inference.baseten.co/v1/models","supports_vision":true,"provider":"baseten","base_model":"decision-models/baseten-joey-2b"},"gpt-6-rosalind":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/pricing","supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","base_model":"gpt-6-rosalind","web_search_billing_unit":"per_request"},"chatgpt/gpt-6-rosalind":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6-rosalind"},"gpt-6-cyber":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"provider":"openai","base_model":"gpt-6-cyber","web_search_billing_unit":"per_request"},"gpt-6-astra-law":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"fedramp_uplift_multiplier":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","base_model":"gpt-6-astra-law","web_search_billing_unit":"per_request"},"chatgpt/gpt-6-astra-law":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6-astra-law"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"max_input_tokens":1048576,"mode":"chat","supports_vision":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"deepseek-v4.1-flash"},"nebius/meta-models/muse-spark-open":{"max_input_tokens":1048576,"mode":"chat","supports_vision":true,"supports_video_input":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"models/muse-spark-open"},"nebius/MiniMaxAI/MiniMax-M3.1-Flash":{"max_input_tokens":1048576,"mode":"chat","supports_vision":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"minimax-m3.1-flash"},"nebius/nvidia/Nemotron-3.5-Super-VL":{"max_input_tokens":1048576,"mode":"chat","supports_vision":true,"supports_video_input":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"nemotron-3.5-super-vl"},"empiriolabs/aplomb-1":{"mode":"chat","max_input_tokens":1000000,"supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/messages","/v1/decisions"],"supports_vision":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/aplomb-1","provider":"empiriolabs","base_model":"aplomb-1"},"empiriolabs/deepseek-v4-1-flash-1017":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-1-flash-1017","supports_prompt_caching":true,"supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-1-flash"},"empiriolabs/deepseek-v4-flash-0731":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-flash-0731","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-flash"},"empiriolabs/deepseek-v4-flash-0731:variant1":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-flash-0731","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-flash-0731-variant1"},"empiriolabs/deepseek-v4-pro":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-pro"},"empiriolabs/deepseek-v4-pro-0813":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-pro-0813","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-pro"},"empiriolabs/deepseek-v4-pro:variant1":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-pro-variant1"},"empiriolabs/deepseek-v4-pro:variant2":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-pro-variant2"},"empiriolabs/deepseek-v4-pro:variant3":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":393216,"supports_system_messages":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/deepseek-v4-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"deepseek-v4-pro-variant3"},"empiriolabs/fugu-max":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/fugu-max","provider":"empiriolabs","base_model":"fugu-max"},"empiriolabs/fugu-ultra-v1-0":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/fugu-ultra-v1-0","provider":"empiriolabs","base_model":"fugu-ultra-v1-0"},"empiriolabs/fugu-ultra-v1-1":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/fugu-ultra-v1-1","provider":"empiriolabs","base_model":"fugu-ultra-v1-1"},"empiriolabs/fugu-ultra-v2-0":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/fugu-ultra-v2-0","provider":"empiriolabs","base_model":"fugu-ultra-v2-0"},"empiriolabs/gemma-4-26b-a4b":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/gemma-4-26b-a4b","provider":"empiriolabs","base_model":"gemma-4-26b-a4b"},"empiriolabs/glm-4-6v-flashx":{"mode":"chat","max_input_tokens":131072,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-4-6v-flashx","provider":"empiriolabs","base_model":"glm-4-6v-flashx"},"empiriolabs/glm-4-7-flashx":{"mode":"chat","max_input_tokens":202752,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-4-7-flashx","provider":"empiriolabs","base_model":"glm-4-7-flashx"},"empiriolabs/glm-5-1":{"mode":"chat","max_input_tokens":202745,"max_output_tokens":128000,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-1","provider":"empiriolabs","base_model":"glm-5-1"},"empiriolabs/glm-5-2":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-2","provider":"empiriolabs","base_model":"glm-5-2"},"empiriolabs/glm-5-2:variant1":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-2","supports_function_calling":true,"provider":"empiriolabs","base_model":"glm-5-2-variant1"},"empiriolabs/glm-5-2:variant2":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-2","supports_function_calling":true,"provider":"empiriolabs","base_model":"glm-5-2-variant2"},"empiriolabs/glm-5-3":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-3","provider":"empiriolabs","base_model":"glm-5-3"},"empiriolabs/glm-5-3-flash":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/glm-5-3-flash","provider":"empiriolabs","base_model":"glm-5-3-flash"},"empiriolabs/kimi-k2-6":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":16000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/kimi-k2-6","supports_function_calling":true,"provider":"empiriolabs","base_model":"kimi-k2-6"},"empiriolabs/kimi-k2-7-code":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/kimi-k2-7-code","supports_function_calling":true,"provider":"empiriolabs","base_model":"kimi-k2-7-code"},"empiriolabs/kimi-k2-7-code-highspeed":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/kimi-k2-7-code-highspeed","supports_function_calling":true,"provider":"empiriolabs","base_model":"kimi-k2-7-code-highspeed"},"empiriolabs/kimi-k2-7-code:variant1":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":16384,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/kimi-k2-7-code","supports_function_calling":true,"provider":"empiriolabs","base_model":"kimi-k2-7-code-variant1"},"empiriolabs/kimi-k3":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/kimi-k3","supports_function_calling":true,"provider":"empiriolabs","base_model":"kimi-k3"},"empiriolabs/kimi-k3-1-pro":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/kimi-k3-1-pro","provider":"empiriolabs","base_model":"kimi-k3-1-pro"},"empiriolabs/mimo-v2-6-flash":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/mimo-v2-6-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"mimo-v2-6-flash"},"empiriolabs/mimo-v2-6-pro":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/mimo-v2-6-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"mimo-v2-6-pro"},"empiriolabs/mimo-v2-6-pro-ultraspeed":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/mimo-v2-6-pro-ultraspeed","supports_function_calling":true,"provider":"empiriolabs","base_model":"mimo-v2-6-pro-ultraspeed"},"empiriolabs/minimax-m2-7":{"mode":"chat","max_input_tokens":204800,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/minimax-m2-7","provider":"empiriolabs","base_model":"minimax-m2-7"},"empiriolabs/minimax-m2-7-highspeed":{"mode":"chat","max_input_tokens":204800,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/minimax-m2-7-highspeed","provider":"empiriolabs","base_model":"minimax-m2-7-highspeed"},"empiriolabs/minimax-m3":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":524288,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/minimax-m3","provider":"empiriolabs","base_model":"minimax-m3"},"empiriolabs/minimax-m3-1-flash":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":524288,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/minimax-m3-1-flash","provider":"empiriolabs","base_model":"minimax-m3-1-flash"},"empiriolabs/minimax-m3:priority":{"mode":"chat","max_input_tokens":1048458,"max_output_tokens":524288,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/minimax-m3","provider":"empiriolabs","base_model":"minimax-m3"},"empiriolabs/mistral-small-4":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/mistral-small-4","supports_function_calling":true,"provider":"empiriolabs","base_model":"mistral-small-4"},"empiriolabs/muse-glimmer-30b":{"mode":"chat","max_input_tokens":131072,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/muse-glimmer-30b","supports_function_calling":true,"provider":"empiriolabs","base_model":"muse-glimmer-30b"},"empiriolabs/muse-spark-1-1":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/muse-spark-1-1","provider":"empiriolabs","base_model":"muse-spark-1-1"},"empiriolabs/muse-spark-1-2":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/muse-spark-1-2","provider":"empiriolabs","base_model":"muse-spark-1-2"},"empiriolabs/muse-spark-1-3":{"mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/muse-spark-1-3","provider":"empiriolabs","base_model":"muse-spark-1-3"},"empiriolabs/nova-lite-1-0":{"mode":"chat","max_input_tokens":317379,"max_output_tokens":10000,"supports_system_messages":true,"supports_vision":true,"supports_prompt_caching":true,"source":"https://empiriolabs.ai/models/nova-lite-1-0","supports_function_calling":true,"provider":"empiriolabs","base_model":"nova-lite-1-0"},"empiriolabs/nova-lite-2":{"mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://empiriolabs.ai/models/nova-lite-2","supports_function_calling":true,"provider":"empiriolabs","base_model":"nova-lite-2"},"empiriolabs/nova-micro-1-0":{"mode":"chat","max_input_tokens":133003,"max_output_tokens":10000,"supports_system_messages":true,"supports_prompt_caching":true,"source":"https://empiriolabs.ai/models/nova-micro-1-0","supports_function_calling":true,"provider":"empiriolabs","base_model":"nova-micro-1-0"},"empiriolabs/nova-pro-1-0":{"mode":"chat","max_input_tokens":317319,"max_output_tokens":10000,"supports_system_messages":true,"supports_vision":true,"source":"https://empiriolabs.ai/models/nova-pro-1-0","supports_function_calling":true,"provider":"empiriolabs","base_model":"nova-pro-1-0"},"empiriolabs/perplexity-advanced-deep-research":{"mode":"chat","max_output_tokens":16000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-advanced-deep-research","provider":"empiriolabs","base_model":"perplexity-advanced"},"empiriolabs/perplexity-deep-research":{"mode":"chat","max_input_tokens":919327,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-deep-research","provider":"empiriolabs","base_model":"perplexity"},"empiriolabs/perplexity-pro-search":{"mode":"chat","max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-pro-search","provider":"empiriolabs","base_model":"perplexity-pro-search"},"empiriolabs/perplexity-sonar":{"mode":"chat","max_input_tokens":919327,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-sonar","provider":"empiriolabs","base_model":"perplexity-sonar"},"empiriolabs/perplexity-sonar-pro":{"mode":"chat","max_input_tokens":919328,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-sonar-pro","provider":"empiriolabs","base_model":"perplexity-sonar-pro"},"empiriolabs/perplexity-sonar-reasoning-pro":{"mode":"chat","max_input_tokens":919330,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"source":"https://empiriolabs.ai/models/perplexity-sonar-reasoning-pro","provider":"empiriolabs","base_model":"perplexity-sonar-reasoning-pro"},"empiriolabs/qwen3-5-122b-a10b":{"mode":"chat","max_input_tokens":258048,"max_output_tokens":64000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-122b-a10b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-122b-a10b"},"empiriolabs/qwen3-5-27b":{"mode":"chat","max_input_tokens":258048,"max_output_tokens":64000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-27b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-27b"},"empiriolabs/qwen3-5-35b-a3b":{"mode":"chat","max_input_tokens":258048,"max_output_tokens":64000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-35b-a3b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-35b-a3b"},"empiriolabs/qwen3-5-397b-a17b":{"mode":"chat","max_input_tokens":258048,"max_output_tokens":64000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-397b-a17b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-397b-a17b"},"empiriolabs/qwen3-5-4b":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-4b","provider":"empiriolabs","base_model":"qwen3-5-4b"},"empiriolabs/qwen3-5-9b":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-9b","provider":"empiriolabs","base_model":"qwen3-5-9b"},"empiriolabs/qwen3-5-flash":{"mode":"chat","max_input_tokens":995904,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-flash"},"empiriolabs/qwen3-5-flash:variant1":{"mode":"chat","max_input_tokens":995904,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-flash-variant1"},"empiriolabs/qwen3-5-omni-flash":{"mode":"chat","max_input_tokens":196608,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/qwen3-5-omni-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-omni-flash"},"empiriolabs/qwen3-5-omni-plus":{"mode":"chat","max_input_tokens":196608,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_web_search":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/qwen3-5-omni-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-omni-plus"},"empiriolabs/qwen3-5-plus":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-plus"},"empiriolabs/qwen3-5-plus:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-5-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-5-plus-variant1"},"empiriolabs/qwen3-6-27b":{"mode":"chat","max_input_tokens":229376,"max_output_tokens":64000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-27b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-27b"},"empiriolabs/qwen3-6-35b-a3b":{"mode":"chat","max_input_tokens":131072,"max_output_tokens":16384,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-35b-a3b","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-35b-a3b"},"empiriolabs/qwen3-6-flash":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-flash"},"empiriolabs/qwen3-6-flash:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-flash-variant1"},"empiriolabs/qwen3-6-plus":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-plus"},"empiriolabs/qwen3-6-plus:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-6-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-6-plus-variant1"},"empiriolabs/qwen3-7-flash":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-flash"},"empiriolabs/qwen3-7-flash:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-flash-variant1"},"empiriolabs/qwen3-7-max":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-max","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-max"},"empiriolabs/qwen3-7-max:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-max","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-max-variant1"},"empiriolabs/qwen3-7-plus":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-plus"},"empiriolabs/qwen3-7-plus:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-7-plus","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-7-plus-variant1"},"empiriolabs/qwen3-8-27b":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-8-27b","provider":"empiriolabs","base_model":"qwen3-8-27b"},"empiriolabs/qwen3-8-flash":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-8-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-8-flash"},"empiriolabs/qwen3-8-max":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-8-max","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-8-max"},"empiriolabs/qwen3-8-max-0902":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-8-max-0902","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-8-max"},"empiriolabs/qwen3-8-max:variant1":{"mode":"chat","max_input_tokens":983616,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/qwen3-8-max","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-8-max-variant1"},"empiriolabs/qwen3-8-omni-flash":{"mode":"chat","max_input_tokens":1006592,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/qwen3-8-omni-flash","supports_function_calling":true,"provider":"empiriolabs","base_model":"qwen3-8-omni-flash"},"empiriolabs/seed-2-0-code":{"mode":"chat","max_input_tokens":229096,"max_output_tokens":128000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"source":"https://empiriolabs.ai/models/seed-2-0-code","supports_function_calling":true,"provider":"empiriolabs","base_model":"seed-2-0-code"},"empiriolabs/seed-2-0-lite":{"mode":"chat","max_input_tokens":229096,"max_output_tokens":128000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/seed-2-0-lite","supports_function_calling":true,"provider":"empiriolabs","base_model":"seed-2-0-lite"},"empiriolabs/seed-2-0-mini":{"mode":"chat","max_input_tokens":229156,"max_output_tokens":128000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/seed-2-0-mini","supports_function_calling":true,"provider":"empiriolabs","base_model":"seed-2-0-mini"},"empiriolabs/seed-2-0-pro":{"mode":"chat","max_input_tokens":229096,"max_output_tokens":128000,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/seed-2-0-pro","supports_function_calling":true,"provider":"empiriolabs","base_model":"seed-2-0-pro"},"empiriolabs/seed-2-1-turbo":{"mode":"chat","max_input_tokens":229093,"max_output_tokens":65536,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/seed-2-1-turbo","supports_function_calling":true,"provider":"empiriolabs","base_model":"seed-2-1-turbo"},"empiriolabs/skylark-embedding-vision":{"mode":"embedding","max_input_tokens":8000,"supports_embedding_image_input":true,"source":"https://empiriolabs.ai/models/skylark-embedding-vision","provider":"empiriolabs","base_model":"skylark-embedding-vision"},"empiriolabs/step-3-5-flash":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/step-3-5-flash","provider":"empiriolabs","base_model":"step-3-5-flash"},"empiriolabs/step-3-5-flash-2603":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"source":"https://empiriolabs.ai/models/step-3-5-flash-2603","provider":"empiriolabs","base_model":"step-3-5-flash"},"empiriolabs/step-3-7-flash":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/step-3-7-flash","provider":"empiriolabs","base_model":"step-3-7-flash"},"empiriolabs/step-5-preview":{"mode":"chat","max_input_tokens":1024000,"max_output_tokens":131072,"supports_system_messages":true,"supports_vision":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"source":"https://empiriolabs.ai/models/step-5-preview","provider":"empiriolabs","base_model":"step-5"},"empiriolabs/stepaudio-2-5-chat":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_web_search":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/stepaudio-2-5-chat","provider":"empiriolabs","base_model":"stepaudio-2-5-chat"},"empiriolabs/stepaudio-3-chat":{"mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"supports_system_messages":true,"supports_reasoning":true,"supports_prompt_caching":true,"supports_function_calling":true,"supports_response_schema":true,"supports_audio_input":true,"source":"https://empiriolabs.ai/models/stepaudio-3-chat","provider":"empiriolabs","base_model":"stepaudio-3-chat"},"empiriolabs/text-embedding-v4":{"mode":"embedding","max_input_tokens":8192,"source":"https://empiriolabs.ai/models/text-embedding-v4","provider":"empiriolabs","base_model":"text-embedding-v4"},"empiriolabs/tongyi-embedding-vision-flash":{"mode":"embedding","max_input_tokens":1024,"supports_embedding_image_input":true,"source":"https://empiriolabs.ai/models/tongyi-embedding-vision-flash","provider":"empiriolabs","base_model":"tongyi-embedding-vision-flash"},"empiriolabs/tongyi-embedding-vision-plus":{"mode":"embedding","max_input_tokens":1024,"supports_embedding_image_input":true,"source":"https://empiriolabs.ai/models/tongyi-embedding-vision-plus","provider":"empiriolabs","base_model":"tongyi-embedding-vision-plus"},"wally/glm-5.3-flash":{"max_input_tokens":1048567,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["minimal","low","medium","high","xhigh","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"glm-5.3-flash"},"wally/deepseek-v4.1-flash":{"max_input_tokens":1048569,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wally","base_model":"deepseek-v4.1-flash"},"wally/kimi-k3":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wally","base_model":"kimi-k3"},"wally/deepseek-v4.1-flash-fast":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wally","base_model":"deepseek-v4.1-flash-fast"},"wally/glm-5.3-fast":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","medium","high","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"glm-5.3-fast"},"wally/deepseek-v4-pro-fast":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","minimal","low","medium","high","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"deepseek-v4-pro-fast"},"wally/deepseek-v4-flash-0731":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","minimal","low","medium","high","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"deepseek-v4-flash"},"wally/minimax-m3":{"max_input_tokens":524287,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://www.runwally.com/pricing","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"minimax-m3"},"wally/deepseek-v4.1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","minimal","low","medium","high","max"],"source":"https://inference.runanywhere.ai/v1/models","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"wally","base_model":"deepseek-v4.1-flash"},"wally/deepseek-v4.1-flash-f2":{"max_input_tokens":1048576,"max_output_tokens":65536,"mode":"chat","reasoning_effort_levels":["none","minimal","low","medium","high","max"],"source":"https://inference.runanywhere.ai/v1/models","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"wally","base_model":"deepseek-v4.1-flash-f2"},"together_ai/moonshotai/Kimi-K3.1-Pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","base_model":"kimi-k3.1-pro"},"gmi/MiniMaxAI/MiniMax-M2":{"max_input_tokens":196608,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"minimax-m2"},"gmi/MiniMaxAI/MiniMax-M2.5":{"max_input_tokens":196608,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"minimax-m2.5"},"gmi/MiniMaxAI/MiniMax-M2.7":{"max_input_tokens":196608,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"minimax-m2.7"},"gmi/MiniMaxAI/MiniMax-M3":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"minimax-m3"},"gmi/MiniMaxAI/MiniMax-M3.1-Flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"minimax-m3.1-flash"},"gmi/Qwen/Qwen3-235B-A22B-FP8":{"max_input_tokens":32768,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-235b-a22b-fp8"},"gmi/Qwen/Qwen3-235B-A22B-Instruct-2507-FP8":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-235b-a22b-instruct-2507-fp8"},"gmi/Qwen/Qwen3-235B-A22B-Thinking-2507-FP8":{"max_input_tokens":262141,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-235b-a22b-thinking-2507-fp8"},"gmi/Qwen/Qwen3-30B-A3B":{"max_input_tokens":40946,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-30b-a3b"},"gmi/Qwen/Qwen3-32B-FP8":{"max_input_tokens":32768,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-32b-fp8"},"gmi/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"max_input_tokens":262128,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-coder-480b-a35b-instruct-fp8"},"gmi/Qwen/Qwen3-Next-80B-A3B-Instruct":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-next-80b-a3b-instruct"},"gmi/Qwen/Qwen3-Next-80B-A3B-Thinking":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3-next-80b-a3b-thinking"},"gmi/Qwen/Qwen3.5-122B-A10B":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.5-122b-a10b"},"gmi/Qwen/Qwen3.5-27B":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.5-27b"},"gmi/Qwen/Qwen3.5-35B-A3B":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.5-35b-a3b"},"gmi/Qwen/Qwen3.5-397B-A17B":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.5-397b-a17b"},"gmi/Qwen/Qwen3.6-35B-A3B":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.6-35b-a3b"},"gmi/Qwen/Qwen3.6-Plus":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.6-plus"},"gmi/Qwen/Qwen3.6-Plus-2026-04-02":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.6-plus"},"gmi/Qwen/Qwen3.7-Max":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.7-max"},"gmi/Qwen/Qwen3.8-27B":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.8-27b"},"gmi/Qwen/Qwen3.8-Flash":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.8-flash"},"gmi/Qwen/Qwen3.8-Max":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.8-max"},"gmi/Qwen/Qwen3.8-Max-0902":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"qwen3.8-max"},"gmi/XiaomiMiMo/MiMo-V2.5":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"mimo-v2.5"},"gmi/XiaomiMiMo/MiMo-V2.5-Pro":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"mimo-v2.5-pro"},"gmi/XiaomiMiMo/MiMo-V2.6-Flash":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"mimo-v2.6-flash"},"gmi/XiaomiMiMo/MiMo-V2.6-Pro":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"mimo-v2.6-pro"},"gmi/deepseek-ai/DeepSeek-Prover-V2-671B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-prover-v2-671b"},"gmi/deepseek-ai/DeepSeek-R1":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1"},"gmi/deepseek-ai/DeepSeek-R1-0528":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1"},"gmi/deepseek-ai/DeepSeek-R1-Distill-Llama-70B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1-distill-llama-70b"},"gmi/deepseek-ai/DeepSeek-R1-Distill-Llama-8B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1-distill-llama-8b"},"gmi/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1-distill-qwen-14b"},"gmi/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1-distill-qwen-32b"},"gmi/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-r1-distill-qwen-7b"},"gmi/deepseek-ai/DeepSeek-V3.1":{"max_input_tokens":163840,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v3.1"},"gmi/deepseek-ai/DeepSeek-V3.1-Terminus":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v3.1-terminus"},"gmi/deepseek-ai/DeepSeek-V3.2-Exp":{"max_input_tokens":163840,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v3.2-exp"},"gmi/deepseek-ai/DeepSeek-V3.2-Speciale":{"max_input_tokens":163840,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v3.2-speciale"},"gmi/deepseek-ai/DeepSeek-V4-Flash":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4-flash"},"gmi/deepseek-ai/DeepSeek-V4-Flash-0731":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4-flash"},"gmi/deepseek-ai/DeepSeek-V4-Pro-0813":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4-pro"},"gmi/deepseek-ai/DeepSeek-V4.1-Flash":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4.1-flash"},"gmi/deepseek-ai/deepseek-v4-flash-vision-exp":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4-flash-vision"},"gmi/deepseek-ai/deepseek-v4.1-flash-1017":{"max_input_tokens":1048575,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"deepseek-v4.1-flash"},"gmi/gmicloud/Ripple-65B-Preview":{"max_input_tokens":1000000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gmicloud/ripple-65b-preview"},"gmi/google/gemini-3.1-flash-lite-preview":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.1-flash-lite"},"gmi/google/gemini-3.1-pro-preview":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.1-pro"},"gmi/google/gemini-3.5-flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.5-flash"},"gmi/google/gemini-3.5-flash-lite":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.5-flash-lite"},"gmi/google/gemini-3.6-flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.6-flash"},"gmi/google/gemini-3.7-flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.7-flash"},"gmi/google/gemini-3.8-flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-3.8-flash"},"gmi/google/gemini-4-argon":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-4-argon"},"gmi/google/gemini-4-carbon":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemini-4-carbon"},"gmi/google/gemma-4-26b-a4b-it":{"max_input_tokens":256000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemma-4-26b-a4b-it"},"gmi/google/gemma-4-31b-it":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gemma-4-31b-it"},"gmi/meta-llama/Llama-3.3-70B-Instruct":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"llama-3.3-70b-instruct"},"gmi/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"llama-4-maverick-17b-128e-instruct-fp8"},"gmi/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"llama-4-scout-17b-16e-instruct"},"gmi/moonshotai/Kimi-K2-Instruct-0905":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k2-instruct"},"gmi/moonshotai/Kimi-K2.5":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k2.5"},"gmi/moonshotai/Kimi-K2.6":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k2.6"},"gmi/moonshotai/kimi-k2.7-code":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k2.7-code"},"gmi/moonshotai/kimi-k2.7-code-highspeed":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k2.7-code-highspeed"},"gmi/moonshotai/kimi-k3":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k3"},"gmi/moonshotai/kimi-k3.1-pro":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"kimi-k3.1-pro"},"gmi/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"max_input_tokens":65536,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"nvidia-nemotron-3.5-lightning-30b-a3b-bf16"},"gmi/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"max_input_tokens":32768,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"nvidia-nemotron-3.5-super-vl-120b-a12b-bf16"},"gmi/nvidia/nemotron-3-ultra-550b-a55b":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"nemotron-3-ultra-550b-a55b"},"gmi/openai/gpt-5.4":{"max_input_tokens":409600,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.4"},"gmi/openai/gpt-5.4-mini":{"max_input_tokens":409600,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.4-mini"},"gmi/openai/gpt-5.4-nano":{"max_input_tokens":409600,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.4-nano"},"gmi/openai/gpt-5.4-pro":{"max_input_tokens":409600,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.4-pro"},"gmi/openai/gpt-5.5":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.5"},"gmi/openai/gpt-5.6-luna":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.6-luna"},"gmi/openai/gpt-5.6-sol":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.6-sol"},"gmi/openai/gpt-5.6-terra":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-5.6-terra"},"gmi/openai/gpt-6-astra":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-6-astra"},"gmi/openai/gpt-6-luna":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-6-luna"},"gmi/openai/gpt-6-sol":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-6-sol"},"gmi/openai/gpt-6.1-sol":{"max_input_tokens":1050000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-6.1-sol"},"gmi/openai/gpt-oss-120b":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-oss-120b"},"gmi/openai/gpt-oss-20b":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"gpt-oss-20b"},"gmi/stepfun-ai/Step-3.7-Flash":{"max_input_tokens":256000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"stepfun-ai/step-3.7-flash"},"gmi/stepfun-ai/Step-5-Preview":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"stepfun-ai/step-5-preview"},"gmi/tencent/HunyuanOCR":{"max_input_tokens":32768,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"hunyuanocr"},"gmi/tencent/Hy3":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"hy3"},"gmi/tencent/hy3-preview":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"hy3"},"gmi/tencent/hy4":{"max_input_tokens":1000000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"hy4"},"gmi/tencent/hy4-preview":{"max_input_tokens":262144,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"hy4"},"gmi/x-ai/grok-4.5":{"max_input_tokens":2000000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"grok-4.5"},"gmi/xai/grok4.6":{"max_input_tokens":2000000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"grok4.6"},"gmi/xai/grok4.7":{"max_input_tokens":500000,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"grok4.7"},"gmi/zai-org/GLM-4.5-Air-FP8":{"max_input_tokens":131076,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-4.5-air-fp8"},"gmi/zai-org/GLM-4.5-FP8":{"max_input_tokens":131072,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-4.5-fp8"},"gmi/zai-org/GLM-4.6":{"max_input_tokens":204800,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-4.6"},"gmi/zai-org/GLM-4.7-Flash":{"max_input_tokens":202752,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-4.7-flash"},"gmi/zai-org/GLM-5-FP8":{"max_input_tokens":202752,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-5-fp8"},"gmi/zai-org/GLM-5.1-FP8":{"max_input_tokens":202752,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-5.1-fp8"},"gmi/zai-org/GLM-5.2-FP8":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-5.2-fp8"},"gmi/zai-org/GLM-5.3":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-5.3"},"gmi/zai-org/GLM-5.3-Flash":{"max_input_tokens":1048576,"mode":"chat","source":"https://ie-api.gmicloud.ai/v1/artifact/get_public_artifacts","provider":"gmi","base_model":"glm-5.3-flash"},"fireworks_ai/accounts/fireworks/models/kimi-k3p1-pro":{"max_input_tokens":1048576,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["low","high","max"],"source":"https://fireworks.ai/models/fireworks/kimi-k3p1-pro","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"fireworks_ai","base_model":"kimi-k3.1-pro","max_tokens":131072,"default_reasoning_effort":"max","thinking_always_on":true,"supported_modalities":["text","image"],"supported_output_modalities":["text"]},"baseten/moonshotai/Kimi-K3.1-Pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://inference.baseten.co/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"baseten","base_model":"kimi-k3.1-pro"},"gpt-image-2.6-sunburst":{"mode":"image_generation","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_pdf_input":true,"supports_vision":true,"provider":"openai","base_model":"gpt-image-2.6-sunburst","web_search_billing_unit":"per_request"},"gpt-image-2.6-flare":{"mode":"image_generation","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/images/generations","/v1/images/edits"],"supports_pdf_input":true,"supports_vision":true,"provider":"openai","base_model":"gpt-image-2.6-flare","web_search_billing_unit":"per_request"},"bedrock/us-gov-east-1/moonshotai.kimi-k3":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-east-1/index.json","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"token_count_framing":64,"token_count_tokenizer":"kimi_k3","provider":"bedrock","base_model":"kimi-k3"},"bedrock/us-gov-west-1/moonshotai.kimi-k3":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","source":"https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonBedrock/current/us-gov-west-1/index.json","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"token_count_framing":64,"token_count_tokenizer":"kimi_k3","provider":"bedrock","base_model":"kimi-k3"},"bedrock_mantle/us-gov-east-1/xai.grok-4.7":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"grok-4.7","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","eu-central-1","eu-central-2","eu-south-1","eu-south-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","ap-east-2","me-central-1","il-central-1","sa-east-1","us-gov-east-1","us-gov-west-1"],"max_tokens":524288000000},"bedrock_mantle/us-gov-west-1/xai.grok-4.7":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"grok-4.7","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","eu-central-1","eu-central-2","eu-south-1","eu-south-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","ap-east-2","me-central-1","il-central-1","sa-east-1","us-gov-east-1","us-gov-west-1"],"max_tokens":524288000000},"bedrock_mantle/xai.grok-4.7":{"max_input_tokens":524288,"max_output_tokens":524288,"mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"grok-4.7","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-north-1","eu-west-1","eu-west-2","eu-west-3","eu-central-1","eu-central-2","eu-south-1","eu-south-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-7","ap-east-2","me-central-1","il-central-1","sa-east-1","us-gov-east-1","us-gov-west-1"],"max_tokens":524288000000},"bedrock_mantle/grok-latest":{"alias_of":"bedrock_mantle/xai.grok-4.7","base_model":"grok"},"together_ai/ricursive-ai/Ripple-65B-Preview":{"max_input_tokens":131072,"max_output_tokens":131072,"mode":"chat","source":"https://api.together.ai/v1/models","supports_prompt_caching":true,"provider":"together_ai","base_model":"ricursive-ai/ripple-65b-preview"},"vertex_ai/gemini-omni-1.2-flash":{"max_input_tokens":131072,"max_output_tokens":260640,"mode":"chat","source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1beta/interactions"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","video"],"supports_audio_input":false,"supports_audio_output":false,"supports_computer_use":false,"supports_function_calling":false,"supports_prompt_caching":false,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":false,"supports_tool_choice":false,"supports_url_context":false,"supports_video_input":true,"supports_vision":true,"supports_web_search":false,"provider":"vertex_ai","base_model":"gemini-omni-1.2-flash"},"gemini/gemini-omni-1.2-flash":{"max_input_tokens":131072,"max_output_tokens":65536,"mode":"chat","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1beta/interactions"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text","video","audio"],"supports_audio_input":false,"supports_audio_output":true,"supports_reasoning":true,"supports_system_messages":false,"supports_video_input":true,"supports_vision":true,"provider":"gemini","base_model":"gemini-omni-1.2-flash"},"databricks/databricks-kimi-k3-1-pro":{"max_input_tokens":1048576,"max_output_tokens":1048576,"metadata":{"notes":"Input/output cost per token is dbu cost * $0.070. Billing reads the per-token dollar fields; the '*_dbu_cost_per_token' fields are the published Databricks rates, kept for reference."},"mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_vision":true,"token_count_framing":100,"provider":"databricks","base_model":"kimi-k3-1-pro"},"moonshotai.kimi-k3.1":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3-1-highspeed.md","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"kimi-k3.1"},"us.moonshotai.kimi-k3.1":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3-1-highspeed.md","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"kimi-k3.1"},"global.moonshotai.kimi-k3.1":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3-1-highspeed.md","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"kimi-k3.1"},"in.moonshotai.kimi-k3.1":{"max_input_tokens":868928,"max_output_tokens":131072,"mode":"chat","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k3-1-highspeed.md","supports_audio_input":false,"supports_bedrock_converse_cache_point":false,"supports_function_calling":true,"supports_none_reasoning_effort":true,"supports_prompt_cache_breakpoint":false,"supports_prompt_caching":true,"supports_reasoning":true,"supports_regex_lookaround":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"provider":"bedrock","base_model":"kimi-k3.1"},"wandb/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"max_input_tokens":1049000,"max_output_tokens":1048576,"mode":"chat","source":"https://www.coreweave.com/products/coreweave-forge/serverless-inference","supports_prompt_caching":true,"supports_reasoning":true,"supports_vision":true,"provider":"wandb","base_model":"deepseek-v4.1-flash"},"azure/cohere-rerank-v5-pro":{"deprecation_date":"2027-02-12","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"rerank-v5-pro"},"azure/cohere-rerank-v5-fast":{"deprecation_date":"2027-02-12","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":32768,"max_output_tokens":32768,"mode":"rerank","source":"https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20'Foundry%20Models'%20and%20armRegionName%20eq%20'eastus'%20and%20priceType%20eq%20'Consumption'","provider":"azure","base_model":"rerank-v5-fast"},"azure/us/gpt-realtime-2.1":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-2.1"},"azure/us/gpt-realtime-2.1-mini":{"deprecation_date":"2027-06-25","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-2.1-mini"},"azure/us/gpt-realtime-1.5":{"deprecation_date":"2027-08-24","deprecation_source":"https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/model-retirement-schedule","max_input_tokens":32000,"max_output_tokens":4096,"mode":"realtime","provider_specific_entry":{"realtime_ga_only":1},"source":"https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_system_messages":true,"supports_tool_choice":true,"provider":"azure","base_model":"gpt-realtime-1.5","supports_prompt_caching":true},"anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"}},"eu.anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_bedrock_native_structured_output_by_default":true,"supports_web_search":true,"supports_safeguards":true,"supports_audio_input":false,"supports_audio_output":false,"token_count_model":{"model":"anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"eu-west-1"}},"global.anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_bedrock_native_structured_output_by_default":true,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"global.anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"}},"us-gov.anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2"},"us.anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000,"tool_use_system_prompt_tokens":346,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_bedrock_native_structured_output_by_default":true,"supports_web_search":true,"supports_safeguards":true,"token_count_model":{"model":"anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"}},"azure/claude-mythos-5-2":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"thinking_always_on":true,"provider":"azure","base_model":"claude-mythos-5-2","max_tokens":128000,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_forced_tool_choice":false,"supports_safeguards":true},"azure/claude-mythos-5-2-fast":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":false,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"thinking_always_on":true,"provider":"azure","base_model":"claude-mythos-5-2-fast"},"azure/claude-fable-5-2-fast":{"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://platform.claude.com/docs/en/about-claude/pricing","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":false,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_web_search":true,"thinking_always_on":true,"provider":"azure","base_model":"claude-fable-5-2-fast"},"bedrock/us-gov-east-1/anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000},"bedrock/us-gov-west-1/anthropic.claude-mythos-5-2":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_bedrock_native_structured_output":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"provider":"bedrock","base_model":"claude-mythos-5-2","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1","us-gov-east-1","us-gov-west-1"],"max_tokens":128000},"github_copilot/claude-fable-5.2":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/chat/completions","/v1/messages"],"supports_adaptive_thinking":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"thinking_always_on":true,"provider":"github_copilot","base_model":"claude-fable-5-2"},"together_ai/elorian-labs/glimpse-1.0":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"together_ai","base_model":"elorian-labs/glimpse-1.0"},"vertex_ai/gemini-4-carbon-cyber-permissive":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_computer_use":false,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-4-carbon-cyber-permissive"},"github_copilot/gemini-4-carbon":{"max_input_tokens":200000,"max_output_tokens":65536,"mode":"chat","source":"https://github.blog/changelog/2026-11-09-gemini-4-carbon-in-github-copilot","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"github_copilot","base_model":"gemini-4-carbon"},"fal_ai/alibaba/happy-horse/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/reference-to-video"},"fal_ai/alibaba/happy-horse/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/text-to-video"},"fal_ai/alibaba/happy-horse/v1.1/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/v1.1/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/v1.1/image-to-video"},"fal_ai/alibaba/happy-horse/v1.1/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/v1.1/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/v1.1/reference-to-video"},"fal_ai/alibaba/happy-horse/v1.1/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/v1.1/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/v1.1/text-to-video"},"fal_ai/alibaba/wan-3.0-prime/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0-prime/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0-prime/image-to-video"},"fal_ai/alibaba/wan-3.0-prime/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0-prime/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0-prime/reference-to-video"},"fal_ai/alibaba/wan-3.0-prime/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0-prime/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0-prime/text-to-video"},"fal_ai/alibaba/wan-3.0/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0/image-to-video"},"fal_ai/alibaba/wan-3.0/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0/reference-to-video"},"fal_ai/alibaba/wan-3.0/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/wan-3.0/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-3.0/text-to-video"},"fal_ai/blackforestlabs/flux-3/draft-enhance":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/draft-enhance","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/draft-enhance"},"fal_ai/blackforestlabs/flux-3/edit-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/edit-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/edit-video"},"fal_ai/blackforestlabs/flux-3/extend-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/extend-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/extend-video"},"fal_ai/blackforestlabs/flux-3/extend-video/draft":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/extend-video/draft","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/extend-video/draft"},"fal_ai/blackforestlabs/flux-3/first-last-frame-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/first-last-frame-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/first-last-frame-to-video"},"fal_ai/blackforestlabs/flux-3/first-last-frame-to-video/draft":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/first-last-frame-to-video/draft","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/first-last-frame-to-video/draft"},"fal_ai/blackforestlabs/flux-3/image-to-video/draft":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/image-to-video/draft","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/image-to-video/draft"},"fal_ai/blackforestlabs/flux-3/keyframes-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/keyframes-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/keyframes-to-video"},"fal_ai/blackforestlabs/flux-3/keyframes-to-video/draft":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/keyframes-to-video/draft","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/keyframes-to-video/draft"},"fal_ai/blackforestlabs/flux-3/klein/base/edit-image/lora":{"metadata":{"notes":"fal bills blackforestlabs/flux-3/klein/base/edit-image/lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/base/edit-image/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/base/edit-image/lora"},"fal_ai/blackforestlabs/flux-3/klein/base/text-to-image/lora":{"metadata":{"notes":"fal bills blackforestlabs/flux-3/klein/base/text-to-image/lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/base/text-to-image/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/base/text-to-image/lora"},"fal_ai/blackforestlabs/flux-3/klein/edit-image/lora":{"metadata":{"notes":"fal bills blackforestlabs/flux-3/klein/edit-image/lora at $0.012 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/edit-image/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/edit-image/lora"},"fal_ai/blackforestlabs/flux-3/klein/text-to-image/lora":{"metadata":{"notes":"fal bills blackforestlabs/flux-3/klein/text-to-image/lora at $0.012 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/text-to-image/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/text-to-image/lora"},"fal_ai/blackforestlabs/flux-3/text-to-video/draft":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/text-to-video/draft","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/text-to-video/draft"},"fal_ai/bria/bria_video_eraser/erase/keypoints":{"mode":"video_generation","source":"https://fal.ai/models/bria/bria_video_eraser/erase/keypoints","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/bria-video-eraser/erase/keypoints"},"fal_ai/bria/bria_video_eraser/erase/mask":{"mode":"video_generation","source":"https://fal.ai/models/bria/bria_video_eraser/erase/mask","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/bria-video-eraser/erase/mask"},"fal_ai/bria/bria_video_eraser/erase/prompt":{"mode":"video_generation","source":"https://fal.ai/models/bria/bria_video_eraser/erase/prompt","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/bria-video-eraser/erase/prompt"},"fal_ai/bria/embed-product":{"mode":"image_generation","source":"https://fal.ai/models/bria/embed-product","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/embed-product"},"fal_ai/bria/extract-object":{"mode":"image_generation","source":"https://fal.ai/models/bria/extract-object","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/extract-object"},"fal_ai/bria/fibo-2/edit":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-2/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-2/edit"},"fal_ai/bria/fibo-bbq-preview/generate":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-bbq-preview/generate","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/fibo-bbq-preview/generate"},"fal_ai/bria/fibo-edit-1.5/edit":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit-1.5/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit-1.5/edit"},"fal_ai/bria/fibo-edit-1.5/expand":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit-1.5/expand","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit-1.5/expand"},"fal_ai/bria/fibo-edit-1.5/product-holding":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit-1.5/product-holding","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit-1.5/product-holding"},"fal_ai/bria/fibo-edit-1.5/virtual-try-on":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit-1.5/virtual-try-on","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit-1.5/virtual-try-on"},"fal_ai/bria/fibo-edit/add_object_by_text":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/add_object_by_text","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/add-object-by-text"},"fal_ai/bria/fibo-edit/blend":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/blend","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/blend"},"fal_ai/bria/fibo-edit/colorize":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/colorize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/colorize"},"fal_ai/bria/fibo-edit/edit":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/edit"},"fal_ai/bria/fibo-edit/erase_by_text":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/erase_by_text","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/erase-by-text"},"fal_ai/bria/fibo-edit/relight":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/relight","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/relight"},"fal_ai/bria/fibo-edit/replace_object_by_text":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/replace_object_by_text","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/replace-object-by-text"},"fal_ai/bria/fibo-edit/reseason":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/reseason","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/reseason"},"fal_ai/bria/fibo-edit/restore":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/restore","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/restore"},"fal_ai/bria/fibo-edit/restyle":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/restyle","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/restyle"},"fal_ai/bria/fibo-edit/rewrite_text":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/rewrite_text","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/rewrite-text"},"fal_ai/bria/fibo-edit/sketch_to_colored_image":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-edit/sketch_to_colored_image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/fibo-edit/sketch-to-colored-image"},"fal_ai/bria/fibo-gen-1.5/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-gen-1.5/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/fibo-gen-1.5/text-to-image"},"fal_ai/bria/fibo-lite/generate":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-lite/generate","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/fibo-lite/generate"},"fal_ai/bria/fibo/generate":{"mode":"image_generation","source":"https://fal.ai/models/bria/fibo/generate","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/fibo/generate"},"fal_ai/bria/genfill/v2":{"metadata":{"notes":"fal bills bria/genfill/v2 at $0.04 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/bria/genfill/v2","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/genfill/v2"},"fal_ai/bria/guided-remove-background":{"mode":"image_generation","source":"https://fal.ai/models/bria/guided-remove-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/guided-remove-background"},"fal_ai/bria/increase-resolution":{"mode":"image_generation","source":"https://fal.ai/models/bria/increase-resolution","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/increase-resolution"},"fal_ai/bria/product-dimensions":{"mode":"image_generation","source":"https://fal.ai/models/bria/product-dimensions","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/product-dimensions"},"fal_ai/bria/replace-background":{"mode":"image_generation","source":"https://fal.ai/models/bria/replace-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/replace-background"},"fal_ai/bria/upscale/creative":{"mode":"image_generation","source":"https://fal.ai/models/bria/upscale/creative","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/upscale/creative"},"fal_ai/bria/video/background-removal":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/background-removal","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/background-removal"},"fal_ai/bria/video/background-removal/green-screen-despill":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/background-removal/green-screen-despill","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/background-removal/green-screen-despill"},"fal_ai/bria/video/background-removal/realtime":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/background-removal/realtime","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/background-removal/realtime"},"fal_ai/bria/video/background-removal/v3":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/background-removal/v3","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/background-removal/v3"},"fal_ai/bria/video/erase/keypoints":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/erase/keypoints","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/erase/keypoints"},"fal_ai/bria/video/erase/mask":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/erase/mask","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/erase/mask"},"fal_ai/bria/video/erase/prompt":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/erase/prompt","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/erase/prompt"},"fal_ai/bria/video/increase-resolution":{"mode":"video_generation","source":"https://fal.ai/models/bria/video/increase-resolution","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bria/video/increase-resolution"},"fal_ai/bytedance/omnihuman/v2":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/omnihuman/v2","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"omnihuman/v2"},"fal_ai/bytedance/seedance-2.0/fast/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/fast/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/fast/image-to-video"},"fal_ai/bytedance/seedance-2.0/fast/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/fast/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/fast/text-to-video"},"fal_ai/bytedance/seedance-2.0/mini/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/mini/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/mini/image-to-video"},"fal_ai/bytedance/seedance-2.0/mini/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 480p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/mini/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/mini/reference-to-video"},"fal_ai/bytedance/seedance-2.0/mini/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/mini/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/mini/text-to-video"},"fal_ai/bytedance/seedance-2.0/us/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/us/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/us/image-to-video"},"fal_ai/bytedance/seedance-2.0/us/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/us/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/us/text-to-video"},"fal_ai/bytedance/seedance-2.5/us/draft/complete":{"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/us/draft/complete","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/us/draft/complete"},"fal_ai/bytedance/seedance-2.5/us/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/us/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/us/image-to-video"},"fal_ai/bytedance/seedance-2.5/us/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/us/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/us/text-to-video"},"fal_ai/bytedance/seedream/v5/flash/edit":{"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/flash/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5/flash/edit"},"fal_ai/bytedance/seedream/v5/flash/layerize":{"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/flash/layerize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5/flash/layerize"},"fal_ai/bytedance/seedream/v5/flash/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/flash/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v5/flash/text-to-image"},"fal_ai/bytedance/seedream/v5/lite/edit":{"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/lite/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5/lite/edit"},"fal_ai/bytedance/seedream/v5/lite/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/lite/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v5/lite/text-to-image"},"fal_ai/clarityai/crystal-upscaler":{"metadata":{"notes":"fal bills clarityai/crystal-upscaler at $0.016 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/clarityai/crystal-upscaler","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"clarityai/crystal-upscaler"},"fal_ai/decart/lucy-2-5/realtime":{"mode":"video_generation","source":"https://fal.ai/models/decart/lucy-2-5/realtime","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"decart/lucy-2-5/realtime"},"fal_ai/decart/lucy-2-6/realtime":{"mode":"video_generation","source":"https://fal.ai/models/decart/lucy-2-6/realtime","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"decart/lucy-2-6/realtime"},"fal_ai/decart/lucy-restyle":{"mode":"video_generation","source":"https://fal.ai/models/decart/lucy-restyle","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"decart/lucy-restyle"},"fal_ai/decart/lucy2-vton/realtime":{"mode":"video_generation","source":"https://fal.ai/models/decart/lucy2-vton/realtime","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"decart/lucy2-vton/realtime"},"fal_ai/fal-ai/ai-avatar/multi":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ai-avatar/multi","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ai-avatar/multi"},"fal_ai/fal-ai/ai-avatar/multi-text":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ai-avatar/multi-text","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ai-avatar/multi-text"},"fal_ai/fal-ai/ai-avatar/single-text":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ai-avatar/single-text","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ai-avatar/single-text"},"fal_ai/fal-ai/bagel":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bagel","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bagel"},"fal_ai/fal-ai/bagel/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bagel/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bagel/edit"},"fal_ai/fal-ai/ben/v2/image":{"metadata":{"notes":"fal bills fal-ai/ben/v2/image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ben/v2/image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ben/v2/image"},"fal_ai/fal-ai/boogu-image":{"metadata":{"notes":"fal bills fal-ai/boogu-image at $0.04 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/boogu-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"boogu-image"},"fal_ai/fal-ai/boogu-image/edit":{"metadata":{"notes":"fal bills fal-ai/boogu-image/edit at $0.04 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/boogu-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"boogu-image/edit"},"fal_ai/fal-ai/bria/background/replace":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/background/replace","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/background/replace"},"fal_ai/fal-ai/bria/eraser":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/eraser","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/eraser"},"fal_ai/fal-ai/bria/expand":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/expand","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/expand"},"fal_ai/fal-ai/bria/genfill":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/genfill","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/genfill"},"fal_ai/fal-ai/bria/product-shot":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/product-shot","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/product-shot"},"fal_ai/fal-ai/bria/reimagine":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/reimagine","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/reimagine"},"fal_ai/fal-ai/bria/text-to-image/base":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/text-to-image/base","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/text-to-image/base"},"fal_ai/fal-ai/bria/text-to-image/fast":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/text-to-image/fast","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/text-to-image/fast"},"fal_ai/fal-ai/bria/text-to-image/hd":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/text-to-image/hd","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/text-to-image/hd"},"fal_ai/fal-ai/bytedance/dreamactor/v2":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/bytedance/dreamactor/v2","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"dreamactor/v2"},"fal_ai/fal-ai/bytedance/omnihuman":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/bytedance/omnihuman","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"omnihuman"},"fal_ai/fal-ai/bytedance/omnihuman/v1.5":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/bytedance/omnihuman/v1.5","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"omnihuman/v1.5"},"fal_ai/fal-ai/bytedance/seedream/v4.5/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bytedance/seedream/v4.5/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v4.5/edit"},"fal_ai/fal-ai/bytedance/seedream/v4.5/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bytedance/seedream/v4.5/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v4.5/text-to-image"},"fal_ai/fal-ai/bytedance/seedream/v4/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bytedance/seedream/v4/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v4/edit"},"fal_ai/fal-ai/bytedance/seedream/v4/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bytedance/seedream/v4/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v4/text-to-image"},"fal_ai/fal-ai/clarity-upscaler":{"metadata":{"notes":"fal bills fal-ai/clarity-upscaler at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/clarity-upscaler","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"clarity-upscaler"},"fal_ai/fal-ai/codeformer":{"metadata":{"notes":"fal bills fal-ai/codeformer at $0.0021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/codeformer","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"codeformer"},"fal_ai/fal-ai/control-light":{"metadata":{"notes":"fal bills fal-ai/control-light at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/control-light","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"control-light"},"fal_ai/fal-ai/davinci-magihuman":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/davinci-magihuman","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"davinci-magihuman"},"fal_ai/fal-ai/ddcolor":{"metadata":{"notes":"fal bills fal-ai/ddcolor at $0.001 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ddcolor","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ddcolor"},"fal_ai/fal-ai/depth-anything-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/depth-anything-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"depth-anything-video"},"fal_ai/fal-ai/echomimic-v3":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/echomimic-v3","supported_endpoints":["/v1/videos"],"supported_modalities":["text","audio"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"echomimic-v3"},"fal_ai/fal-ai/ernie-image":{"metadata":{"notes":"fal bills fal-ai/ernie-image at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ernie-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ernie-image"},"fal_ai/fal-ai/ernie-image/lora":{"metadata":{"notes":"fal bills fal-ai/ernie-image/lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ernie-image/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ernie-image/lora"},"fal_ai/fal-ai/ernie-image/lora/turbo":{"metadata":{"notes":"fal bills fal-ai/ernie-image/lora/turbo at $0.015 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ernie-image/lora/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ernie-image/lora/turbo"},"fal_ai/fal-ai/ernie-image/turbo":{"metadata":{"notes":"fal bills fal-ai/ernie-image/turbo at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ernie-image/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ernie-image/turbo"},"fal_ai/fal-ai/fashn/tryon/v1.5":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/fashn/tryon/v1.5","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"fashn/tryon/v1.5"},"fal_ai/fal-ai/fashn/tryon/v1.6":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/fashn/tryon/v1.6","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"fashn/tryon/v1.6"},"fal_ai/fal-ai/feynobg":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/feynobg","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"feynobg"},"fal_ai/fal-ai/ffmpeg-api/compose":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ffmpeg-api/compose","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ffmpeg-api/compose"},"fal_ai/fal-ai/ffmpeg-api/merge-audio-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ffmpeg-api/merge-audio-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ffmpeg-api/merge-audio-video"},"fal_ai/fal-ai/finegrain-eraser":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/finegrain-eraser","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"finegrain-eraser"},"fal_ai/fal-ai/finegrain-eraser/bbox":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/finegrain-eraser/bbox","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"finegrain-eraser/bbox"},"fal_ai/fal-ai/finegrain-eraser/mask":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/finegrain-eraser/mask","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"finegrain-eraser/mask"},"fal_ai/fal-ai/firered-image-edit-v1.1":{"metadata":{"notes":"fal bills fal-ai/firered-image-edit-v1.1 at $0.0325 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/firered-image-edit-v1.1","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"firered-image-edit"},"fal_ai/fal-ai/flashhead":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/flashhead","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"flashhead"},"fal_ai/fal-ai/flux-1/dev":{"metadata":{"notes":"fal bills fal-ai/flux-1/dev at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/dev","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-1/dev"},"fal_ai/fal-ai/flux-1/dev/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-1/dev/image-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/dev/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/dev/image-to-image"},"fal_ai/fal-ai/flux-1/dev/redux":{"metadata":{"notes":"fal bills fal-ai/flux-1/dev/redux at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/dev/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/dev/redux"},"fal_ai/fal-ai/flux-1/krea":{"metadata":{"notes":"fal bills fal-ai/flux-1/krea at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/krea","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-1/krea"},"fal_ai/fal-ai/flux-1/krea/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-1/krea/image-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/krea/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/krea/image-to-image"},"fal_ai/fal-ai/flux-1/krea/redux":{"metadata":{"notes":"fal bills fal-ai/flux-1/krea/redux at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/krea/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/krea/redux"},"fal_ai/fal-ai/flux-1/schnell":{"metadata":{"notes":"fal bills fal-ai/flux-1/schnell at $0.003 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/schnell","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-1/schnell"},"fal_ai/fal-ai/flux-1/schnell/redux":{"metadata":{"notes":"fal bills fal-ai/flux-1/schnell/redux at $0.003 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/schnell/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/schnell/redux"},"fal_ai/fal-ai/flux-1/srpo":{"metadata":{"notes":"fal bills fal-ai/flux-1/srpo at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/srpo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-1/srpo"},"fal_ai/fal-ai/flux-1/srpo/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-1/srpo/image-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-1/srpo/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-1/srpo/image-to-image"},"fal_ai/fal-ai/flux-2":{"metadata":{"notes":"fal bills fal-ai/flux-2 at $0.012 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2"},"fal_ai/fal-ai/flux-2-flex":{"metadata":{"notes":"fal bills fal-ai/flux-2-flex at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-flex","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-flex"},"fal_ai/fal-ai/flux-2-flex/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2-flex/edit at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-flex/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-flex/edit"},"fal_ai/fal-ai/flux-2-lora-gallery/add-background":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/add-background at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/add-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-lora-gallery/add-background"},"fal_ai/fal-ai/flux-2-lora-gallery/apartment-staging":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/apartment-staging at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/apartment-staging","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-lora-gallery/apartment-staging"},"fal_ai/fal-ai/flux-2-lora-gallery/ballpoint-pen-sketch":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/ballpoint-pen-sketch at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/ballpoint-pen-sketch","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/ballpoint-pen-sketch"},"fal_ai/fal-ai/flux-2-lora-gallery/digital-comic-art":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/digital-comic-art at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/digital-comic-art","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/digital-comic-art"},"fal_ai/fal-ai/flux-2-lora-gallery/face-to-full-portrait":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/face-to-full-portrait at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/face-to-full-portrait","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-lora-gallery/face-to-full-portrait"},"fal_ai/fal-ai/flux-2-lora-gallery/hdr-style":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/hdr-style at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/hdr-style","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/hdr-style"},"fal_ai/fal-ai/flux-2-lora-gallery/multiple-angles":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/multiple-angles at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/multiple-angles","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-lora-gallery/multiple-angles"},"fal_ai/fal-ai/flux-2-lora-gallery/realism":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/realism at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/realism","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/realism"},"fal_ai/fal-ai/flux-2-lora-gallery/satellite-view-style":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/satellite-view-style at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/satellite-view-style","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/satellite-view-style"},"fal_ai/fal-ai/flux-2-lora-gallery/sepia-vintage":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/sepia-vintage at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/sepia-vintage","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-lora-gallery/sepia-vintage"},"fal_ai/fal-ai/flux-2-lora-gallery/virtual-tryon":{"metadata":{"notes":"fal bills fal-ai/flux-2-lora-gallery/virtual-tryon at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-lora-gallery/virtual-tryon","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-lora-gallery/virtual-tryon"},"fal_ai/fal-ai/flux-2/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/edit at $0.012 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/edit"},"fal_ai/fal-ai/flux-2/flash":{"metadata":{"notes":"fal bills fal-ai/flux-2/flash at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/flash","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/flash"},"fal_ai/fal-ai/flux-2/flash/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/flash/edit at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/flash/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/flash/edit"},"fal_ai/fal-ai/flux-2/klein/4b":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/4b"},"fal_ai/fal-ai/flux-2/klein/4b/base":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/base at $0.009 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/base","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/4b/base"},"fal_ai/fal-ai/flux-2/klein/4b/base/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/base/edit at $0.009 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/base/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/4b/base/edit"},"fal_ai/fal-ai/flux-2/klein/4b/base/edit/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/base/edit/lora at $0.016 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/base/edit/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/4b/base/edit/lora"},"fal_ai/fal-ai/flux-2/klein/4b/base/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/base/lora at $0.016 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/base/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/4b/base/lora"},"fal_ai/fal-ai/flux-2/klein/4b/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/edit at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/4b/edit"},"fal_ai/fal-ai/flux-2/klein/4b/edit/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/edit/lora at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/edit/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/4b/edit/lora"},"fal_ai/fal-ai/flux-2/klein/4b/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/4b/lora at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/4b/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/4b/lora"},"fal_ai/fal-ai/flux-2/klein/9b":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b at $0.006 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/9b"},"fal_ai/fal-ai/flux-2/klein/9b/base":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/base at $0.011 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/base","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/9b/base"},"fal_ai/fal-ai/flux-2/klein/9b/base/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/base/edit at $0.011 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/base/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/9b/base/edit"},"fal_ai/fal-ai/flux-2/klein/9b/base/edit/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/base/edit/lora at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/base/edit/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/9b/base/edit/lora"},"fal_ai/fal-ai/flux-2/klein/9b/base/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/base/lora at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/base/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/9b/base/lora"},"fal_ai/fal-ai/flux-2/klein/9b/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/edit at $0.011 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/9b/edit"},"fal_ai/fal-ai/flux-2/klein/9b/edit/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/edit/lora at $0.015 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/edit/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/klein/9b/edit/lora"},"fal_ai/fal-ai/flux-2/klein/9b/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/klein/9b/lora at $0.015 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/klein/9b/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/klein/9b/lora"},"fal_ai/fal-ai/flux-2/lora":{"metadata":{"notes":"fal bills fal-ai/flux-2/lora at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/lora"},"fal_ai/fal-ai/flux-2/lora/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/lora/edit at $0.021 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/lora/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/lora/edit"},"fal_ai/fal-ai/flux-2/turbo":{"metadata":{"notes":"fal bills fal-ai/flux-2/turbo at $0.008 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2/turbo"},"fal_ai/fal-ai/flux-2/turbo/edit":{"metadata":{"notes":"fal bills fal-ai/flux-2/turbo/edit at $0.008 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries (input megapixels are billed too and are not modeled)"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2/turbo/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2/turbo/edit"},"fal_ai/fal-ai/flux-general":{"metadata":{"notes":"fal bills fal-ai/flux-general at $0.075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-general","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-general"},"fal_ai/fal-ai/flux-general/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-general/image-to-image at $0.075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-general/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-general/image-to-image"},"fal_ai/fal-ai/flux-general/inpainting":{"metadata":{"notes":"fal bills fal-ai/flux-general/inpainting at $0.075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-general/inpainting","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-general/inpainting"},"fal_ai/fal-ai/flux-kontext-lora":{"metadata":{"notes":"fal bills fal-ai/flux-kontext-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-kontext-lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-kontext-lora"},"fal_ai/fal-ai/flux-kontext-lora/inpaint":{"metadata":{"notes":"fal bills fal-ai/flux-kontext-lora/inpaint at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-kontext-lora/inpaint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-kontext-lora/inpaint"},"fal_ai/fal-ai/flux-kontext-lora/text-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-kontext-lora/text-to-image at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-kontext-lora/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-kontext-lora/text-to-image"},"fal_ai/fal-ai/flux-kontext/dev":{"metadata":{"notes":"fal bills fal-ai/flux-kontext/dev at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-kontext/dev","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-kontext/dev"},"fal_ai/fal-ai/flux-krea-lora":{"metadata":{"notes":"fal bills fal-ai/flux-krea-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-krea-lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-krea-lora"},"fal_ai/fal-ai/flux-krea-lora/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-krea-lora/image-to-image at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-krea-lora/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-krea-lora/image-to-image"},"fal_ai/fal-ai/flux-krea-lora/inpainting":{"metadata":{"notes":"fal bills fal-ai/flux-krea-lora/inpainting at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-krea-lora/inpainting","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-krea-lora/inpainting"},"fal_ai/fal-ai/flux-krea-lora/stream":{"metadata":{"notes":"fal bills fal-ai/flux-krea-lora/stream at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-krea-lora/stream","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-krea-lora/stream"},"fal_ai/fal-ai/flux-lora":{"metadata":{"notes":"fal bills fal-ai/flux-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-lora"},"fal_ai/fal-ai/flux-lora-canny":{"metadata":{"notes":"fal bills fal-ai/flux-lora-canny at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora-canny","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-lora-canny"},"fal_ai/fal-ai/flux-lora-fill":{"metadata":{"notes":"fal bills fal-ai/flux-lora-fill at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora-fill","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-lora-fill"},"fal_ai/fal-ai/flux-lora/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux-lora/image-to-image at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-lora/image-to-image"},"fal_ai/fal-ai/flux-lora/inpainting":{"metadata":{"notes":"fal bills fal-ai/flux-lora/inpainting at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora/inpainting","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-lora/inpainting"},"fal_ai/fal-ai/flux-lora/stream":{"metadata":{"notes":"fal bills fal-ai/flux-lora/stream at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-lora/stream","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-lora/stream"},"fal_ai/fal-ai/flux-pro/kontext":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/kontext"},"fal_ai/fal-ai/flux-pro/kontext/max":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext/max","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/kontext/max"},"fal_ai/fal-ai/flux-pro/kontext/max/multi":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext/max/multi","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/kontext/max/multi"},"fal_ai/fal-ai/flux-pro/kontext/max/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext/max/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-pro/kontext/max/text-to-image"},"fal_ai/fal-ai/flux-pro/kontext/multi":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext/multi","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/kontext/multi"},"fal_ai/fal-ai/flux-pro/kontext/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/kontext/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-pro/kontext/text-to-image"},"fal_ai/fal-ai/flux-pro/v1.1-ultra-finetuned":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1.1-ultra-finetuned","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-pro/v1.1-ultra-finetuned"},"fal_ai/fal-ai/flux-pro/v1.1-ultra/redux":{"metadata":{"notes":"fal bills fal-ai/flux-pro/v1.1-ultra/redux at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1.1-ultra/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1.1-ultra/redux"},"fal_ai/fal-ai/flux-pro/v1.1/redux":{"metadata":{"notes":"fal bills fal-ai/flux-pro/v1.1/redux at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1.1/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1.1/redux"},"fal_ai/fal-ai/flux-pro/v1/fill":{"metadata":{"notes":"fal bills fal-ai/flux-pro/v1/fill at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1/fill","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1/fill"},"fal_ai/fal-ai/flux-pro/v1/fill-finetuned":{"metadata":{"notes":"fal bills fal-ai/flux-pro/v1/fill-finetuned at $0.06 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1/fill-finetuned","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1/fill-finetuned"},"fal_ai/fal-ai/flux-pulid":{"metadata":{"notes":"fal bills fal-ai/flux-pulid at $0.0333 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pulid","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pulid"},"fal_ai/fal-ai/flux-vision-upscaler":{"metadata":{"notes":"fal bills fal-ai/flux-vision-upscaler at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-vision-upscaler","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-vision-upscaler"},"fal_ai/fal-ai/flux/dev/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux/dev/image-to-image at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/dev/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/dev/image-to-image"},"fal_ai/fal-ai/flux/dev/redux":{"metadata":{"notes":"fal bills fal-ai/flux/dev/redux at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/dev/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/dev/redux"},"fal_ai/fal-ai/flux/krea":{"metadata":{"notes":"fal bills fal-ai/flux/krea at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/krea","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux/krea"},"fal_ai/fal-ai/flux/krea/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux/krea/image-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/krea/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/krea/image-to-image"},"fal_ai/fal-ai/flux/krea/redux":{"metadata":{"notes":"fal bills fal-ai/flux/krea/redux at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/krea/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/krea/redux"},"fal_ai/fal-ai/flux/schnell/redux":{"metadata":{"notes":"fal bills fal-ai/flux/schnell/redux at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/schnell/redux","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/schnell/redux"},"fal_ai/fal-ai/flux/srpo":{"metadata":{"notes":"fal bills fal-ai/flux/srpo at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/srpo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux/srpo"},"fal_ai/fal-ai/flux/srpo/image-to-image":{"metadata":{"notes":"fal bills fal-ai/flux/srpo/image-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux/srpo/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux/srpo/image-to-image"},"fal_ai/fal-ai/framepack":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/framepack","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"framepack"},"fal_ai/fal-ai/framepack/f1":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/framepack/f1","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"framepack/f1"},"fal_ai/fal-ai/framepack/flf2v":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/framepack/flf2v","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"framepack/flf2v"},"fal_ai/fal-ai/gemini-25-flash-image/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-25-flash-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gemini-25-flash-image/edit"},"fal_ai/fal-ai/gemini-3-pro-image-preview":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-3-pro-image-preview","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"gemini-3-pro-image"},"fal_ai/fal-ai/gemini-3-pro-image-preview/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-3-pro-image-preview/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gemini-3-pro-image-preview/edit"},"fal_ai/fal-ai/gemini-3.1-flash-image-preview":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-3.1-flash-image-preview","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"gemini-3.1-flash-image"},"fal_ai/fal-ai/gemini-3.1-flash-image-preview/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/gemini-3.1-flash-image-preview/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"gemini-3.1-flash-image-preview/edit"},"fal_ai/fal-ai/glm-image":{"metadata":{"notes":"fal bills fal-ai/glm-image at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/glm-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"glm-image"},"fal_ai/fal-ai/glm-image/image-to-image":{"metadata":{"notes":"fal bills fal-ai/glm-image/image-to-image at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/glm-image/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"glm-image/image-to-image"},"fal_ai/fal-ai/heygen/avatar3/digital-twin":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/avatar3/digital-twin","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar3/digital-twin"},"fal_ai/fal-ai/heygen/avatar4/digital-twin":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/avatar4/digital-twin","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar4/digital-twin"},"fal_ai/fal-ai/heygen/avatar4/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/avatar4/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar4/image-to-video"},"fal_ai/fal-ai/heygen/avatar5/digital-twin":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/avatar5/digital-twin","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar5/digital-twin"},"fal_ai/fal-ai/heygen/v2/translate/precision":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v2/translate/precision","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v2/translate/precision"},"fal_ai/fal-ai/heygen/v2/translate/speed":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v2/translate/speed","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v2/translate/speed"},"fal_ai/fal-ai/heygen/v2/video-agent":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v2/video-agent","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v2/video-agent"},"fal_ai/fal-ai/heygen/v3/filler-word-removal":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v3/filler-word-removal","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v3/filler-word-removal"},"fal_ai/fal-ai/heygen/v3/lipsync/precision":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v3/lipsync/precision","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v3/lipsync/precision"},"fal_ai/fal-ai/heygen/v3/lipsync/speed":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v3/lipsync/speed","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v3/lipsync/speed"},"fal_ai/fal-ai/heygen/v3/video-agent":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/heygen/v3/video-agent","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/v3/video-agent"},"fal_ai/fal-ai/hidream-i1-fast":{"metadata":{"notes":"fal bills fal-ai/hidream-i1-fast at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-i1-fast","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hidream-i1-fast"},"fal_ai/fal-ai/hidream-i1-full":{"metadata":{"notes":"fal bills fal-ai/hidream-i1-full at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-i1-full","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hidream-i1-full"},"fal_ai/fal-ai/hidream-i1-full/image-to-image":{"metadata":{"notes":"fal bills fal-ai/hidream-i1-full/image-to-image at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-i1-full/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"hidream-i1-full/image-to-image"},"fal_ai/fal-ai/hidream-o1-image":{"metadata":{"notes":"fal bills fal-ai/hidream-o1-image at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-o1-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hidream-o1-image"},"fal_ai/fal-ai/hidream-o1-image/dev":{"metadata":{"notes":"fal bills fal-ai/hidream-o1-image/dev at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-o1-image/dev","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hidream-o1-image/dev"},"fal_ai/fal-ai/hidream-o1-image/dev/edit":{"metadata":{"notes":"fal bills fal-ai/hidream-o1-image/dev/edit at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-o1-image/dev/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"hidream-o1-image/dev/edit"},"fal_ai/fal-ai/hidream-o1-image/edit":{"metadata":{"notes":"fal bills fal-ai/hidream-o1-image/edit at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hidream-o1-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"hidream-o1-image/edit"},"fal_ai/fal-ai/hunyuan_world":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hunyuan_world","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"hunyuan-world"},"fal_ai/fal-ai/hunyuan-image/v2.1/text-to-image":{"metadata":{"notes":"fal bills fal-ai/hunyuan-image/v2.1/text-to-image at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hunyuan-image/v2.1/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hunyuan-image/v2.1/text-to-image"},"fal_ai/fal-ai/hunyuan-image/v3/instruct/edit":{"metadata":{"notes":"fal bills fal-ai/hunyuan-image/v3/instruct/edit at $0.09 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hunyuan-image/v3/instruct/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"hunyuan-image/v3/instruct/edit"},"fal_ai/fal-ai/hunyuan-image/v3/instruct/text-to-image":{"metadata":{"notes":"fal bills fal-ai/hunyuan-image/v3/instruct/text-to-image at $0.09 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hunyuan-image/v3/instruct/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hunyuan-image/v3/instruct/text-to-image"},"fal_ai/fal-ai/hunyuan-image/v3/text-to-image":{"metadata":{"notes":"fal bills fal-ai/hunyuan-image/v3/text-to-image at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/hunyuan-image/v3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"hunyuan-image/v3/text-to-image"},"fal_ai/fal-ai/hunyuan-video-v1.5/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/hunyuan-video-v1.5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hunyuan-video-v1.5/image-to-video"},"fal_ai/fal-ai/hunyuan-video-v1.5/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/hunyuan-video-v1.5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hunyuan-video-v1.5/text-to-video"},"fal_ai/fal-ai/iclight-v2":{"metadata":{"notes":"fal bills fal-ai/iclight-v2 at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/iclight-v2","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"iclight-v2"},"fal_ai/fal-ai/id-v2v":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/id-v2v","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"id-v2v"},"fal_ai/fal-ai/id-v2v/relight":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/id-v2v/relight","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"id-v2v/relight"},"fal_ai/fal-ai/ideogram/character":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/character","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/character"},"fal_ai/fal-ai/ideogram/character/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/character/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/character/edit"},"fal_ai/fal-ai/ideogram/character/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/character/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/character/remix"},"fal_ai/fal-ai/ideogram/object-removal":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/object-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/object-removal"},"fal_ai/fal-ai/ideogram/remove-background":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/remove-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/remove-background"},"fal_ai/fal-ai/ideogram/upscale":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/upscale","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/upscale"},"fal_ai/fal-ai/ideogram/v2":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v2"},"fal_ai/fal-ai/ideogram/v2/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2/edit"},"fal_ai/fal-ai/ideogram/v2/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2/remix"},"fal_ai/fal-ai/ideogram/v2/turbo":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v2/turbo"},"fal_ai/fal-ai/ideogram/v2/turbo/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2/turbo/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2/turbo/edit"},"fal_ai/fal-ai/ideogram/v2/turbo/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2/turbo/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2/turbo/remix"},"fal_ai/fal-ai/ideogram/v2a":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2a","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v2a"},"fal_ai/fal-ai/ideogram/v2a/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2a/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2a/remix"},"fal_ai/fal-ai/ideogram/v2a/turbo":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2a/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v2a/turbo"},"fal_ai/fal-ai/ideogram/v2a/turbo/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v2a/turbo/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v2a/turbo/remix"},"fal_ai/fal-ai/ideogram/v3/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v3/edit"},"fal_ai/fal-ai/ideogram/v3/generate-transparent":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/generate-transparent","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v3/generate-transparent"},"fal_ai/fal-ai/ideogram/v3/layerize-text":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/layerize-text","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v3/layerize-text"},"fal_ai/fal-ai/ideogram/v3/reframe":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/reframe","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v3/reframe"},"fal_ai/fal-ai/ideogram/v3/remix":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/remix","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v3/remix"},"fal_ai/fal-ai/ideogram/v3/replace-background":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/ideogram/v3/replace-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v3/replace-background"},"fal_ai/fal-ai/image-apps-v2/age-modify":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/age-modify","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/age-modify"},"fal_ai/fal-ai/image-apps-v2/city-teleport":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/city-teleport","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/city-teleport"},"fal_ai/fal-ai/image-apps-v2/expression-change":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/expression-change","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/expression-change"},"fal_ai/fal-ai/image-apps-v2/hair-change":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/hair-change","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/hair-change"},"fal_ai/fal-ai/image-apps-v2/headshot-photo":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/headshot-photo","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/headshot-photo"},"fal_ai/fal-ai/image-apps-v2/makeup-application":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/makeup-application","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/makeup-application"},"fal_ai/fal-ai/image-apps-v2/object-removal":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/object-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/object-removal"},"fal_ai/fal-ai/image-apps-v2/outpaint":{"metadata":{"notes":"fal bills fal-ai/image-apps-v2/outpaint at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/outpaint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/outpaint"},"fal_ai/fal-ai/image-apps-v2/perspective":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/perspective","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/perspective"},"fal_ai/fal-ai/image-apps-v2/photo-restoration":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/photo-restoration","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/photo-restoration"},"fal_ai/fal-ai/image-apps-v2/photography-effects":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/photography-effects","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/photography-effects"},"fal_ai/fal-ai/image-apps-v2/portrait-enhance":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/portrait-enhance","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/portrait-enhance"},"fal_ai/fal-ai/image-apps-v2/product-holding":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/product-holding","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/product-holding"},"fal_ai/fal-ai/image-apps-v2/product-photography":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/product-photography","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/product-photography"},"fal_ai/fal-ai/image-apps-v2/relighting":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/relighting","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/relighting"},"fal_ai/fal-ai/image-apps-v2/style-transfer":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/style-transfer","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/style-transfer"},"fal_ai/fal-ai/image-apps-v2/texture-transform":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/texture-transform","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/texture-transform"},"fal_ai/fal-ai/image-apps-v2/virtual-try-on":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-apps-v2/virtual-try-on","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-apps-v2/virtual-try-on"},"fal_ai/fal-ai/image-editing/age-progression":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/age-progression","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/age-progression"},"fal_ai/fal-ai/image-editing/baby-version":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/baby-version","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/baby-version"},"fal_ai/fal-ai/image-editing/background-change":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/background-change","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/background-change"},"fal_ai/fal-ai/image-editing/broccoli-haircut":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/broccoli-haircut","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/broccoli-haircut"},"fal_ai/fal-ai/image-editing/cartoonify":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/cartoonify","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/cartoonify"},"fal_ai/fal-ai/image-editing/color-correction":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/color-correction","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/color-correction"},"fal_ai/fal-ai/image-editing/expression-change":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/expression-change","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/expression-change"},"fal_ai/fal-ai/image-editing/face-enhancement":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/face-enhancement","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/face-enhancement"},"fal_ai/fal-ai/image-editing/hair-change":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/hair-change","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/hair-change"},"fal_ai/fal-ai/image-editing/object-removal":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/object-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/object-removal"},"fal_ai/fal-ai/image-editing/photo-restoration":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/photo-restoration","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/photo-restoration"},"fal_ai/fal-ai/image-editing/plushie-style":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/plushie-style","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/plushie-style"},"fal_ai/fal-ai/image-editing/professional-photo":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/professional-photo","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/professional-photo"},"fal_ai/fal-ai/image-editing/realism":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/realism","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/realism"},"fal_ai/fal-ai/image-editing/reframe":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/reframe","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/reframe"},"fal_ai/fal-ai/image-editing/retouch":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/retouch","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/retouch"},"fal_ai/fal-ai/image-editing/scene-composition":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/scene-composition","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/scene-composition"},"fal_ai/fal-ai/image-editing/style-transfer":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/style-transfer","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/style-transfer"},"fal_ai/fal-ai/image-editing/text-removal":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/text-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/text-removal"},"fal_ai/fal-ai/image-editing/time-of-day":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/time-of-day","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/time-of-day"},"fal_ai/fal-ai/image-editing/weather-effect":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/weather-effect","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/weather-effect"},"fal_ai/fal-ai/image-editing/wojak-style":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/wojak-style","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/wojak-style"},"fal_ai/fal-ai/image-editing/youtube-thumbnails":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image-editing/youtube-thumbnails","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-editing/youtube-thumbnails"},"fal_ai/fal-ai/image2svg":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/image2svg","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image2svg"},"fal_ai/fal-ai/instant-character":{"metadata":{"notes":"fal bills fal-ai/instant-character at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/instant-character","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"instant-character"},"fal_ai/fal-ai/kling-image/o1":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/kling-image/o1","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"kling-image/o1"},"fal_ai/fal-ai/kling-image/o3/image-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/kling-image/o3/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"kling-image/o3/image-to-image"},"fal_ai/fal-ai/kling-image/o3/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/kling-image/o3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"kling-image/o3/text-to-image"},"fal_ai/fal-ai/kling-image/v3/image-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/kling-image/v3/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"kling-image/v3/image-to-image"},"fal_ai/fal-ai/kling-image/v3/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/kling-image/v3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"kling-image/v3/text-to-image"},"fal_ai/fal-ai/kling-video/ai-avatar/v2/pro":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/ai-avatar/v2/pro","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/ai-avatar/v2/pro"},"fal_ai/fal-ai/kling-video/ai-avatar/v2/standard":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/ai-avatar/v2/standard","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/ai-avatar/v2/standard"},"fal_ai/fal-ai/kling-video/o1/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/image-to-video"},"fal_ai/fal-ai/kling-video/o1/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/reference-to-video"},"fal_ai/fal-ai/kling-video/o1/standard/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/standard/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/standard/image-to-video"},"fal_ai/fal-ai/kling-video/o1/standard/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/standard/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/standard/reference-to-video"},"fal_ai/fal-ai/kling-video/o1/standard/video-to-video/edit":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/standard/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/standard/video-to-video/edit"},"fal_ai/fal-ai/kling-video/o1/standard/video-to-video/reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/standard/video-to-video/reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/standard/video-to-video/reference"},"fal_ai/fal-ai/kling-video/o1/video-to-video/edit":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/video-to-video/edit"},"fal_ai/fal-ai/kling-video/o1/video-to-video/reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o1/video-to-video/reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o1/video-to-video/reference"},"fal_ai/fal-ai/kling-video/o3/4k/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/4k/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/4k/image-to-video"},"fal_ai/fal-ai/kling-video/o3/4k/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/4k/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/4k/reference-to-video"},"fal_ai/fal-ai/kling-video/o3/4k/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/4k/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/4k/text-to-video"},"fal_ai/fal-ai/kling-video/o3/4k/video-to-video/edit":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/4k/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/4k/video-to-video/edit"},"fal_ai/fal-ai/kling-video/o3/4k/video-to-video/reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/4k/video-to-video/reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/4k/video-to-video/reference"},"fal_ai/fal-ai/kling-video/o3/pro/video-to-video/edit":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/pro/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/pro/video-to-video/edit"},"fal_ai/fal-ai/kling-video/o3/pro/video-to-video/reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/pro/video-to-video/reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/pro/video-to-video/reference"},"fal_ai/fal-ai/kling-video/o3/standard/video-to-video/edit":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/standard/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/standard/video-to-video/edit"},"fal_ai/fal-ai/kling-video/o3/standard/video-to-video/reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/o3/standard/video-to-video/reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/o3/standard/video-to-video/reference"},"fal_ai/fal-ai/kling-video/v1.5/pro/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1.5/pro/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1.5/pro/effects"},"fal_ai/fal-ai/kling-video/v1.6/pro/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1.6/pro/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1.6/pro/effects"},"fal_ai/fal-ai/kling-video/v1.6/standard/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1.6/standard/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1.6/standard/effects"},"fal_ai/fal-ai/kling-video/v1/pro/ai-avatar":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1/pro/ai-avatar","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1/pro/ai-avatar"},"fal_ai/fal-ai/kling-video/v1/standard/ai-avatar":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1/standard/ai-avatar","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1/standard/ai-avatar"},"fal_ai/fal-ai/kling-video/v1/standard/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v1/standard/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v1/standard/effects"},"fal_ai/fal-ai/kling-video/v2.6/pro/motion-control":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.6/pro/motion-control","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.6/pro/motion-control"},"fal_ai/fal-ai/kling-video/v2.6/standard/motion-control":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.6/standard/motion-control","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.6/standard/motion-control"},"fal_ai/fal-ai/kling-video/v3/4k/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/4k/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/4k/image-to-video"},"fal_ai/fal-ai/kling-video/v3/4k/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/4k/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/4k/text-to-video"},"fal_ai/fal-ai/kling-video/v3/pro/motion-control":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/pro/motion-control","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/pro/motion-control"},"fal_ai/fal-ai/kling-video/v3/standard/motion-control":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/standard/motion-control","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/standard/motion-control"},"fal_ai/fal-ai/kling-video/v3/turbo/pro/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/turbo/pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/turbo/pro/image-to-video"},"fal_ai/fal-ai/kling-video/v3/turbo/pro/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/turbo/pro/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/turbo/pro/text-to-video"},"fal_ai/fal-ai/kling-video/v3/turbo/standard/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/turbo/standard/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/turbo/standard/image-to-video"},"fal_ai/fal-ai/kling-video/v3/turbo/standard/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v3/turbo/standard/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v3/turbo/standard/text-to-video"},"fal_ai/fal-ai/krea-2/turbo":{"metadata":{"notes":"fal bills fal-ai/krea-2/turbo at $0.008 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/krea-2/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"krea-2/turbo"},"fal_ai/fal-ai/krea-2/turbo/lora":{"metadata":{"notes":"fal bills fal-ai/krea-2/turbo/lora at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/krea-2/turbo/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"krea-2/turbo/lora"},"fal_ai/fal-ai/krea-2/turbo/style":{"metadata":{"notes":"fal bills fal-ai/krea-2/turbo/style at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/krea-2/turbo/style","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"krea-2/turbo/style"},"fal_ai/fal-ai/krea-wan-14b/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/krea-wan-14b/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"krea-wan-14b/text-to-video"},"fal_ai/fal-ai/krea-wan-14b/video-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/krea-wan-14b/video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"krea-wan-14b/video-to-video"},"fal_ai/fal-ai/latentsync":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/latentsync","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"latentsync"},"fal_ai/fal-ai/leffa/pose-transfer":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/leffa/pose-transfer","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"leffa/pose-transfer"},"fal_ai/fal-ai/leffa/virtual-tryon":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/leffa/virtual-tryon","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"leffa/virtual-tryon"},"fal_ai/fal-ai/lightx/recamera":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/lightx/recamera","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightx/recamera"},"fal_ai/fal-ai/lightx/relight":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/lightx/relight","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightx/relight"},"fal_ai/fal-ai/longcat-image":{"metadata":{"notes":"fal bills fal-ai/longcat-image at $0.13 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/longcat-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"longcat-image"},"fal_ai/fal-ai/longcat-image/edit":{"metadata":{"notes":"fal bills fal-ai/longcat-image/edit at $0.15 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/longcat-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"longcat-image/edit"},"fal_ai/fal-ai/longcat-video/distilled/image-to-video/480p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/distilled/image-to-video/480p","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/distilled/image-to-video/480p"},"fal_ai/fal-ai/longcat-video/distilled/image-to-video/720p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/distilled/image-to-video/720p","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/distilled/image-to-video/720p"},"fal_ai/fal-ai/longcat-video/distilled/text-to-video/480p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/distilled/text-to-video/480p","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/distilled/text-to-video/480p"},"fal_ai/fal-ai/longcat-video/distilled/text-to-video/720p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/distilled/text-to-video/720p","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/distilled/text-to-video/720p"},"fal_ai/fal-ai/longcat-video/image-to-video/480p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/image-to-video/480p","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/image-to-video/480p"},"fal_ai/fal-ai/longcat-video/image-to-video/720p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/image-to-video/720p","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/image-to-video/720p"},"fal_ai/fal-ai/longcat-video/text-to-video/480p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/text-to-video/480p","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/text-to-video/480p"},"fal_ai/fal-ai/longcat-video/text-to-video/720p":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/longcat-video/text-to-video/720p","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"longcat-video/text-to-video/720p"},"fal_ai/fal-ai/ltx-2.3/audio-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/audio-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","audio"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/audio-to-video"},"fal_ai/fal-ai/ltx-2.3/extend-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/extend-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/extend-video"},"fal_ai/fal-ai/ltx-2.3/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.16, 2160p $0.32"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/image-to-video"},"fal_ai/fal-ai/ltx-2.3/image-to-video/fast":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.12, 2160p $0.24"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/image-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/image-to-video/fast"},"fal_ai/fal-ai/ltx-2.3/reframe":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/reframe","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/reframe"},"fal_ai/fal-ai/ltx-2.3/retake-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/retake-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/retake-video"},"fal_ai/fal-ai/ltx-2.3/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.16, 2160p $0.32"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/text-to-video"},"fal_ai/fal-ai/ltx-2.3/text-to-video/fast":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.12, 2160p $0.24"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-2.3/text-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-2.3/text-to-video/fast"},"fal_ai/fal-ai/ltxv-13b-098-distilled":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltxv-13b-098-distilled","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltxv-13b-098-distilled"},"fal_ai/fal-ai/ltxv-13b-098-distilled/extend":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltxv-13b-098-distilled/extend","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltxv-13b-098-distilled/extend"},"fal_ai/fal-ai/ltxv-13b-098-distilled/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltxv-13b-098-distilled/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltxv-13b-098-distilled/image-to-video"},"fal_ai/fal-ai/ltxv-13b-098-distilled/multiconditioning":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltxv-13b-098-distilled/multiconditioning","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltxv-13b-098-distilled/multiconditioning"},"fal_ai/fal-ai/lumina-image/v2":{"metadata":{"notes":"fal bills fal-ai/lumina-image/v2 at $0.075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/lumina-image/v2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"lumina-image/v2"},"fal_ai/fal-ai/marigold-v2":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/marigold-v2","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"marigold-v2"},"fal_ai/fal-ai/minimax/hailuo-02-fast/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 512p rate; tiers without a catalog field: 512p $0.017"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-02-fast/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-02-fast/image-to-video"},"fal_ai/fal-ai/minimax/hailuo-02/pro/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-02/pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-02/pro/image-to-video"},"fal_ai/fal-ai/minimax/hailuo-02/pro/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-02/pro/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-02/pro/text-to-video"},"fal_ai/fal-ai/minimax/hailuo-02/standard/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-02/standard/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-02/standard/text-to-video"},"fal_ai/fal-ai/minimax/image-01/subject-reference":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/minimax/image-01/subject-reference","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"image-01/subject-reference"},"fal_ai/fal-ai/mmaudio-v2":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/mmaudio-v2","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"mmaudio-v2"},"fal_ai/fal-ai/moss-image":{"metadata":{"notes":"fal bills fal-ai/moss-image at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/moss-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"moss-image"},"fal_ai/fal-ai/nafnet/deblur":{"metadata":{"notes":"fal bills fal-ai/nafnet/deblur at $0.0225 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nafnet/deblur","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nafnet/deblur"},"fal_ai/fal-ai/nafnet/denoise":{"metadata":{"notes":"fal bills fal-ai/nafnet/denoise at $0.0225 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nafnet/denoise","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nafnet/denoise"},"fal_ai/fal-ai/nano-banana-2/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana-2/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nano-banana-2/edit"},"fal_ai/fal-ai/nano-banana-pro/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana-pro/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nano-banana-pro/edit"},"fal_ai/fal-ai/nano-banana/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nano-banana/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"nano-banana/edit"},"fal_ai/fal-ai/nucleus-image":{"metadata":{"notes":"fal bills fal-ai/nucleus-image at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/nucleus-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"nucleus-image"},"fal_ai/fal-ai/omnigen-v2":{"metadata":{"notes":"fal bills fal-ai/omnigen-v2 at $0.15 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/omnigen-v2","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"omnigen-v2"},"fal_ai/fal-ai/one-to-all-animation/1.3b":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 480p rate; tiers without a catalog field: 580p $0.045"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/one-to-all-animation/1.3b","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"one-to-all-animation/1.3b"},"fal_ai/fal-ai/one-to-all-animation/14b":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 480p rate; tiers without a catalog field: 580p $0.09"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/one-to-all-animation/14b","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"one-to-all-animation/14b"},"fal_ai/fal-ai/patina":{"metadata":{"notes":"fal bills fal-ai/patina at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/patina","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"patina"},"fal_ai/fal-ai/phota/enhance":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/phota/enhance","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"phota/enhance"},"fal_ai/fal-ai/post-processing/blur":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/blur","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/blur"},"fal_ai/fal-ai/post-processing/chromatic-aberration":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/chromatic-aberration","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/chromatic-aberration"},"fal_ai/fal-ai/post-processing/color-correction":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/color-correction","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/color-correction"},"fal_ai/fal-ai/post-processing/color-tint":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/color-tint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/color-tint"},"fal_ai/fal-ai/post-processing/desaturate":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/desaturate","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/desaturate"},"fal_ai/fal-ai/post-processing/dissolve":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/dissolve","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/dissolve"},"fal_ai/fal-ai/post-processing/dodge-burn":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/dodge-burn","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/dodge-burn"},"fal_ai/fal-ai/post-processing/grain":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/grain","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/grain"},"fal_ai/fal-ai/post-processing/parabolize":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/parabolize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/parabolize"},"fal_ai/fal-ai/post-processing/sharpen":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/sharpen","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/sharpen"},"fal_ai/fal-ai/post-processing/solarize":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/solarize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/solarize"},"fal_ai/fal-ai/post-processing/vignette":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/post-processing/vignette","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"post-processing/vignette"},"fal_ai/fal-ai/qwen-image":{"metadata":{"notes":"fal bills fal-ai/qwen-image at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"qwen-image"},"fal_ai/fal-ai/qwen-image-2512":{"metadata":{"notes":"fal bills fal-ai/qwen-image-2512 at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-2512","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"qwen-image"},"fal_ai/fal-ai/qwen-image-2512/lora":{"metadata":{"notes":"fal bills fal-ai/qwen-image-2512/lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-2512/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"qwen-image-2512/lora"},"fal_ai/fal-ai/qwen-image-edit-2509":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509 at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit"},"fal_ai/fal-ai/qwen-image-edit-2509-lora":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/add-background":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/add-background at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/add-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/add-background"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/face-to-full-portrait":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/face-to-full-portrait at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/face-to-full-portrait","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/face-to-full-portrait"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/group-photo":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/group-photo at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/group-photo","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/group-photo"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/integrate-product":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/integrate-product at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/integrate-product","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/integrate-product"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/lighting-restoration":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/lighting-restoration at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/lighting-restoration","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/lighting-restoration"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/multiple-angles":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/multiple-angles at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/multiple-angles","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/multiple-angles"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/next-scene":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/next-scene at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/next-scene","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/next-scene"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/remove-element":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/remove-element at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/remove-element","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/remove-element"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/remove-lighting":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/remove-lighting at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/remove-lighting","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/remove-lighting"},"fal_ai/fal-ai/qwen-image-edit-2509-lora-gallery/shirt-design":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2509-lora-gallery/shirt-design at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2509-lora-gallery/shirt-design","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2509-lora-gallery/shirt-design"},"fal_ai/fal-ai/qwen-image-edit-2511":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2511 at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2511","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit"},"fal_ai/fal-ai/qwen-image-edit-2511-multiple-angles":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2511-multiple-angles at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2511-multiple-angles","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2511-multiple-angles"},"fal_ai/fal-ai/qwen-image-edit-2511/lora":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-2511/lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-2511/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-2511/lora"},"fal_ai/fal-ai/qwen-image-edit-lora":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-lora"},"fal_ai/fal-ai/qwen-image-edit-plus":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus"},"fal_ai/fal-ai/qwen-image-edit-plus-lora":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/add-background":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/add-background at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/add-background","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/add-background"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/face-to-full-portrait":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/face-to-full-portrait at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/face-to-full-portrait","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/face-to-full-portrait"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/group-photo":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/group-photo at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/group-photo","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/group-photo"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/integrate-product":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/integrate-product at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/integrate-product","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/integrate-product"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/lighting-restoration":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/lighting-restoration at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/lighting-restoration","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/lighting-restoration"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/multiple-angles":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/multiple-angles at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/multiple-angles","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/multiple-angles"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/next-scene":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/next-scene at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/next-scene","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/next-scene"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/remove-element":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/remove-element at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/remove-element","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/remove-element"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/remove-lighting":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/remove-lighting at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/remove-lighting","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/remove-lighting"},"fal_ai/fal-ai/qwen-image-edit-plus-lora-gallery/shirt-design":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit-plus-lora-gallery/shirt-design at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit-plus-lora-gallery/shirt-design","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit-plus-lora-gallery/shirt-design"},"fal_ai/fal-ai/qwen-image-edit/image-to-image":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit/image-to-image at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit/image-to-image"},"fal_ai/fal-ai/qwen-image-edit/inpaint":{"metadata":{"notes":"fal bills fal-ai/qwen-image-edit/inpaint at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-edit/inpaint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-edit/inpaint"},"fal_ai/fal-ai/qwen-image-layered":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-layered","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-layered"},"fal_ai/fal-ai/qwen-image-layered/lora":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image-layered/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image-layered/lora"},"fal_ai/fal-ai/qwen-image/image-to-image":{"metadata":{"notes":"fal bills fal-ai/qwen-image/image-to-image at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/qwen-image/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"qwen-image/image-to-image"},"fal_ai/fal-ai/recraft/upscale/creative":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/upscale/creative","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"upscale/creative"},"fal_ai/fal-ai/recraft/upscale/crisp":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/upscale/crisp","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"upscale/crisp"},"fal_ai/fal-ai/recraft/v4.1/pro/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/pro/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/pro/text-to-image"},"fal_ai/fal-ai/recraft/v4.1/pro/text-to-vector":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/pro/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/pro/text-to-vector"},"fal_ai/fal-ai/recraft/v4.1/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/text-to-image"},"fal_ai/fal-ai/recraft/v4.1/text-to-vector":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/text-to-vector"},"fal_ai/fal-ai/recraft/v4.1/utility/pro/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/utility/pro/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/utility/pro/text-to-image"},"fal_ai/fal-ai/recraft/v4.1/utility/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4.1/utility/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/utility/text-to-image"},"fal_ai/fal-ai/recraft/v4/pro/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4/pro/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/pro/text-to-image"},"fal_ai/fal-ai/recraft/v4/pro/text-to-vector":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4/pro/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/pro/text-to-vector"},"fal_ai/fal-ai/recraft/v4/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/text-to-image"},"fal_ai/fal-ai/recraft/v4/text-to-vector":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v4/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/text-to-vector"},"fal_ai/fal-ai/recraft/vectorize":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/vectorize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"vectorize"},"fal_ai/fal-ai/sana":{"metadata":{"notes":"fal bills fal-ai/sana at $0.001 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/sana","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"sana"},"fal_ai/fal-ai/sana/sprint":{"metadata":{"notes":"fal bills fal-ai/sana/sprint at $0.0025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/sana/sprint","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"sana/sprint"},"fal_ai/fal-ai/sana/v1.5/1.6b":{"metadata":{"notes":"fal bills fal-ai/sana/v1.5/1.6b at $0.0075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/sana/v1.5/1.6b","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"sana/v1.5/1.6b"},"fal_ai/fal-ai/sana/v1.5/4.8b":{"metadata":{"notes":"fal bills fal-ai/sana/v1.5/4.8b at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/sana/v1.5/4.8b","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"sana/v1.5/4.8b"},"fal_ai/fal-ai/seedvr/upscale/image":{"metadata":{"notes":"fal bills fal-ai/seedvr/upscale/image at $0.001 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/seedvr/upscale/image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedvr/upscale/image"},"fal_ai/fal-ai/seedvr/upscale/image/seamless":{"metadata":{"notes":"fal bills fal-ai/seedvr/upscale/image/seamless at $0.0025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/seedvr/upscale/image/seamless","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedvr/upscale/image/seamless"},"fal_ai/fal-ai/stable-diffusion-v3-medium":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/stable-diffusion-v3-medium","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"stable-diffusion-v3-medium"},"fal_ai/fal-ai/stable-diffusion-v35-large":{"metadata":{"notes":"fal bills fal-ai/stable-diffusion-v35-large at $0.065 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/stable-diffusion-v35-large","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"stable-diffusion-v35-large"},"fal_ai/fal-ai/stepx-edit2":{"metadata":{"notes":"fal bills fal-ai/stepx-edit2 at $0.2 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/stepx-edit2","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"stepx-edit2"},"fal_ai/fal-ai/telestyle-v2":{"metadata":{"notes":"fal bills fal-ai/telestyle-v2 at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/telestyle-v2","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"telestyle-v2"},"fal_ai/fal-ai/uso":{"metadata":{"notes":"fal bills fal-ai/uso at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/uso","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"uso"},"fal_ai/fal-ai/vidu/q2/reference-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/vidu/q2/reference-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"vidu/q2/reference-to-image"},"fal_ai/fal-ai/vidu/q2/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/vidu/q2/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"vidu/q2/text-to-image"},"fal_ai/fal-ai/vidu/reference-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/vidu/reference-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"vidu/reference-to-image"},"fal_ai/fal-ai/wan-22-vace-fun-a14b/depth":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.075"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-22-vace-fun-a14b/depth","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-22-vace-fun-a14b/depth"},"fal_ai/fal-ai/auto-caption":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/auto-caption","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"auto-caption"},"fal_ai/fal-ai/hunyuan-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/hunyuan-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hunyuan-video"},"fal_ai/fal-ai/hunyuan-video-image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/hunyuan-video-image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hunyuan-video-image-to-video"},"fal_ai/fal-ai/infinity-star/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/infinity-star/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"infinity-star/text-to-video"},"fal_ai/fal-ai/ltx-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-video"},"fal_ai/fal-ai/ltx-video-13b-distilled":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-video-13b-distilled","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-video-13b-distilled"},"fal_ai/fal-ai/ltx-video-13b-distilled/extend":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-video-13b-distilled/extend","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-video-13b-distilled/extend"},"fal_ai/fal-ai/ltx-video-13b-distilled/multiconditioning":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-video-13b-distilled/multiconditioning","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-video-13b-distilled/multiconditioning"},"fal_ai/fal-ai/ltx-video/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ltx-video/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ltx-video/image-to-video"},"fal_ai/fal-ai/minimax/hailuo-2.3-fast/pro/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-2.3-fast/pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-2.3-fast/pro/image-to-video"},"fal_ai/fal-ai/minimax/video-01":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01"},"fal_ai/fal-ai/minimax/video-01-director":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01-director","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01-director"},"fal_ai/fal-ai/minimax/video-01-director/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01-director/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01-director/image-to-video"},"fal_ai/fal-ai/minimax/video-01-live":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01-live","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01-live"},"fal_ai/fal-ai/minimax/video-01-live/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01-live/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01-live/image-to-video"},"fal_ai/fal-ai/minimax/video-01-subject-reference":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01-subject-reference","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01-subject-reference"},"fal_ai/fal-ai/minimax/video-01/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/video-01/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"video-01/image-to-video"},"fal_ai/fal-ai/ovi":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ovi","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ovi"},"fal_ai/fal-ai/ovi/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/ovi/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"ovi/image-to-video"},"fal_ai/fal-ai/wan-i2v-lora":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-i2v-lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-i2v-lora"},"fal_ai/fal-ai/wan-pro/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-pro/image-to-video"},"fal_ai/fal-ai/wan-t2v-lora":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-t2v-lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-t2v-lora"},"fal_ai/fal-ai/wan/v2.2-5b/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-5b/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-5b/image-to-video"},"fal_ai/fal-ai/wan/v2.2-5b/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-5b/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-5b/text-to-video"},"fal_ai/moonvalley/marey/motion-transfer":{"mode":"video_generation","source":"https://fal.ai/models/moonvalley/marey/motion-transfer","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"moonvalley/marey/motion-transfer"},"fal_ai/moonvalley/marey/pose-transfer":{"mode":"video_generation","source":"https://fal.ai/models/moonvalley/marey/pose-transfer","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"moonvalley/marey/pose-transfer"},"fal_ai/mureka/api/generate/lyrics-video":{"mode":"video_generation","source":"https://fal.ai/models/mureka/api/generate/lyrics-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"mureka/api/generate/lyrics-video"},"fal_ai/fal-ai/wan-22-vace-fun-a14b/inpainting":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.075"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-22-vace-fun-a14b/inpainting","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-22-vace-fun-a14b/inpainting"},"fal_ai/fal-ai/wan-22-vace-fun-a14b/outpainting":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.075"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-22-vace-fun-a14b/outpainting","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-22-vace-fun-a14b/outpainting"},"fal_ai/fal-ai/wan-22-vace-fun-a14b/reframe":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.075"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-22-vace-fun-a14b/reframe","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-22-vace-fun-a14b/reframe"},"fal_ai/fal-ai/wan-25-preview/image-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan-25-preview/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"wan-25-preview/image-to-image"},"fal_ai/fal-ai/wan-25-preview/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-25-preview/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-25-preview/image-to-video"},"fal_ai/fal-ai/wan-25-preview/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan-25-preview/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan-25-preview/text-to-image"},"fal_ai/fal-ai/wan-25-preview/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-25-preview/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-25-preview/text-to-video"},"fal_ai/fal-ai/wan-vace-14b":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b"},"fal_ai/fal-ai/wan-vace-14b/depth":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b/depth","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b/depth"},"fal_ai/fal-ai/wan-vace-14b/inpainting":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b/inpainting","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b/inpainting"},"fal_ai/fal-ai/wan-vace-14b/outpainting":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b/outpainting","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b/outpainting"},"fal_ai/fal-ai/wan-vace-14b/pose":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b/pose","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b/pose"},"fal_ai/fal-ai/wan-vace-14b/reframe":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-14b/reframe","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-14b/reframe"},"fal_ai/fal-ai/wan-vace-apps/video-edit":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p); tiers without a catalog field: 580p $0.075"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-apps/video-edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-apps/video-edit"},"fal_ai/fal-ai/wan/v2.2-14b/speech-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 480p rate; tiers without a catalog field: 580p $0.15"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-14b/speech-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","audio"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-14b/speech-to-video"},"fal_ai/fal-ai/wan/v2.2-5b/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-5b/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.2-5b/text-to-image"},"fal_ai/fal-ai/wan/v2.2-a14b/image-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"wan/v2.2-a14b/image-to-image"},"fal_ai/fal-ai/wan/v2.2-a14b/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/image-to-video"},"fal_ai/fal-ai/wan/v2.2-a14b/image-to-video/lora":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/image-to-video/lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/image-to-video/lora"},"fal_ai/fal-ai/wan/v2.2-a14b/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/text-to-image"},"fal_ai/fal-ai/wan/v2.2-a14b/text-to-image/lora":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/text-to-image/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/text-to-image/lora"},"fal_ai/fal-ai/wan/v2.2-a14b/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/text-to-video"},"fal_ai/fal-ai/wan/v2.2-a14b/text-to-video/lora":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/text-to-video/lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/text-to-video/lora"},"fal_ai/fal-ai/wan/v2.2-a14b/video-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 580p $0.06"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-a14b/video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-a14b/video-to-video"},"fal_ai/fal-ai/wan/v2.7/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"wan/v2.7/edit"},"fal_ai/fal-ai/wan/v2.7/edit-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/edit-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.7/edit-video"},"fal_ai/fal-ai/wan/v2.7/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.7/image-to-video"},"fal_ai/fal-ai/wan/v2.7/pro/edit":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/pro/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"wan/v2.7/pro/edit"},"fal_ai/fal-ai/wan/v2.7/pro/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/pro/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.7/pro/text-to-image"},"fal_ai/fal-ai/wan/v2.7/reference-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.7/reference-to-video"},"fal_ai/fal-ai/wan/v2.7/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.7/text-to-image"},"fal_ai/fal-ai/wan/v2.7/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.7/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.7/text-to-video"},"fal_ai/fal-ai/z-image/base":{"metadata":{"notes":"fal bills fal-ai/z-image/base at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/base","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/base"},"fal_ai/fal-ai/z-image/base/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/base/lora at $0.012 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/base/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/base/lora"},"fal_ai/fal-ai/z-image/turbo":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/turbo"},"fal_ai/fal-ai/z-image/turbo/controlnet":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/controlnet at $0.0065 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/controlnet","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/controlnet"},"fal_ai/fal-ai/z-image/turbo/controlnet/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/controlnet/lora at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/controlnet/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/controlnet/lora"},"fal_ai/fal-ai/z-image/turbo/image-to-image":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/image-to-image at $0.005 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/image-to-image"},"fal_ai/fal-ai/z-image/turbo/image-to-image/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/image-to-image/lora at $0.0085 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/image-to-image/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/image-to-image/lora"},"fal_ai/fal-ai/z-image/turbo/inpaint":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/inpaint at $0.01 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/inpaint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/inpaint"},"fal_ai/fal-ai/z-image/turbo/inpaint/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/inpaint/lora at $0.0115 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/inpaint/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"z-image/turbo/inpaint/lora"},"fal_ai/fal-ai/z-image/turbo/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/lora at $0.0085 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/turbo/lora"},"fal_ai/fal-ai/z-image/turbo/tiling":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/tiling at $0.02 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/tiling","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/turbo/tiling"},"fal_ai/fal-ai/z-image/turbo/tiling/lora":{"metadata":{"notes":"fal bills fal-ai/z-image/turbo/tiling/lora at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/z-image/turbo/tiling/lora","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"z-image/turbo/tiling/lora"},"fal_ai/google/gemini-omni-flash/v1.1/edit":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 360p $0.03"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.1/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.1/edit"},"fal_ai/google/gemini-omni-flash/v1.1/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 360p $0.03"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.1/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.1/image-to-video"},"fal_ai/google/gemini-omni-flash/v1.1/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 360p $0.03"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.1/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.1/reference-to-video"},"fal_ai/google/gemini-omni-flash/v1.1/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 360p $0.03"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.1/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.1/text-to-video"},"fal_ai/kling/v4/flash/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/flash/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/flash/image-to-video"},"fal_ai/kling/v4/flash/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/flash/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/flash/text-to-video"},"fal_ai/kling/v4/flash/video-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/flash/video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/flash/video-to-video"},"fal_ai/kling/v4/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/image-to-video"},"fal_ai/kling/v4/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/text-to-video"},"fal_ai/kling/v4/video-to-video/edit":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/video-to-video/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/video-to-video/edit"},"fal_ai/krea/v3/edit":{"metadata":{"notes":"fal bills krea/v3/edit at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/krea/v3/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"krea/v3/edit"},"fal_ai/krea/v3/text-to-image":{"metadata":{"notes":"fal bills krea/v3/text-to-image at $0.025 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/krea/v3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"krea/v3/text-to-image"},"fal_ai/luma/agent/uni-1/v1/max":{"mode":"image_generation","source":"https://fal.ai/models/luma/agent/uni-1/v1/max","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"luma/agent/uni-1/v1/max"},"fal_ai/luma/agent/uni-1/v1/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/luma/agent/uni-1/v1/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"luma/agent/uni-1/v1/text-to-image"},"fal_ai/meta/muse-image/edit":{"mode":"image_generation","source":"https://fal.ai/models/meta/muse-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"muse-image/edit"},"fal_ai/meta/muse-image/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/meta/muse-image/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"muse-image/text-to-image"},"fal_ai/minimax/h3-max-turbo/extend-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max-turbo/extend-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max-turbo/extend-video"},"fal_ai/minimax/h3-max-turbo/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max-turbo/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max-turbo/reference-to-video"},"fal_ai/minimax/h3-max/extend-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/extend-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/extend-video"},"fal_ai/minimax/h3-max/insert-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/insert-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/insert-video"},"fal_ai/minimax/h3-max/lip-sync/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/lip-sync/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/lip-sync/image-to-video"},"fal_ai/minimax/h3/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 2k rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/image-to-video"},"fal_ai/minimax/h3/image-to-video/lora":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 2k rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/image-to-video/lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/image-to-video/lora"},"fal_ai/minimax/h3/reference-to-video/lora":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 2k rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/reference-to-video/lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/reference-to-video/lora"},"fal_ai/minimax/h3/text-to-video/lora":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 2k rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3/text-to-video/lora","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3/text-to-video/lora"},"fal_ai/nvidia/cosmos-3-super/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/nvidia/cosmos-3-super/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"cosmos-3-super/image-to-video"},"fal_ai/pixelcut/background-removal":{"mode":"image_generation","source":"https://fal.ai/models/pixelcut/background-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"pixelcut/background-removal"},"fal_ai/pixelcut/looping-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/pixelcut/looping-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixelcut/looping-video"},"fal_ai/pixelcut/product-photo":{"mode":"image_generation","source":"https://fal.ai/models/pixelcut/product-photo","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"pixelcut/product-photo"},"fal_ai/pixverse/music-video/vibemv":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/pixverse/music-video/vibemv","supported_endpoints":["/v1/videos"],"supported_modalities":["text","audio"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/music-video/vibemv"},"fal_ai/recraft/v4.1/flash/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/recraft/v4.1/flash/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4.1/flash/text-to-image"},"fal_ai/rundiffusion-fal/juggernaut-flux/base":{"metadata":{"notes":"fal bills rundiffusion-fal/juggernaut-flux/base at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/juggernaut-flux/base","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"rundiffusion-fal/juggernaut-flux/base"},"fal_ai/rundiffusion-fal/juggernaut-flux/base/image-to-image":{"metadata":{"notes":"fal bills rundiffusion-fal/juggernaut-flux/base/image-to-image at $0.035 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/juggernaut-flux/base/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"rundiffusion-fal/juggernaut-flux/base/image-to-image"},"fal_ai/rundiffusion-fal/juggernaut-flux/lightning":{"metadata":{"notes":"fal bills rundiffusion-fal/juggernaut-flux/lightning at $0.006 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/juggernaut-flux/lightning","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"rundiffusion-fal/juggernaut-flux/lightning"},"fal_ai/rundiffusion-fal/juggernaut-flux/pro":{"metadata":{"notes":"fal bills rundiffusion-fal/juggernaut-flux/pro at $0.055 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/juggernaut-flux/pro","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"rundiffusion-fal/juggernaut-flux/pro"},"fal_ai/rundiffusion-fal/juggernaut-flux/pro/image-to-image":{"metadata":{"notes":"fal bills rundiffusion-fal/juggernaut-flux/pro/image-to-image at $0.055 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/juggernaut-flux/pro/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"rundiffusion-fal/juggernaut-flux/pro/image-to-image"},"fal_ai/rundiffusion-fal/rundiffusion-photo-flux":{"metadata":{"notes":"fal bills rundiffusion-fal/rundiffusion-photo-flux at $0.045 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/rundiffusion-fal/rundiffusion-photo-flux","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"rundiffusion-fal/rundiffusion-photo-flux"},"fal_ai/smoretalk-ai/rembg-enhance":{"mode":"image_generation","source":"https://fal.ai/models/smoretalk-ai/rembg-enhance","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"smoretalk-ai/rembg-enhance"},"fal_ai/sonilo/v1.1/video-to-video-music":{"mode":"video_generation","source":"https://fal.ai/models/sonilo/v1.1/video-to-video-music","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"sonilo/v1.1/video-to-video-music"},"fal_ai/sonilo/v1.1/video-to-video-sound-effects":{"mode":"video_generation","source":"https://fal.ai/models/sonilo/v1.1/video-to-video-sound-effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"sonilo/v1.1/video-to-video-sound-effects"},"fal_ai/tbc/mischief-1/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/tbc/mischief-1/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"tbc/mischief-1/image-to-video"},"fal_ai/veed/fabric-1.0":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p)"},"mode":"video_generation","source":"https://fal.ai/models/veed/fabric-1.0","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"veed/fabric-1.0"},"fal_ai/veed/fabric-1.0/fast":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p)"},"mode":"video_generation","source":"https://fal.ai/models/veed/fabric-1.0/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"veed/fabric-1.0/fast"},"fal_ai/veed/fabric-1.0/text":{"metadata":{"comment":"per second of generated video by resolution; the page's resolution default is auto or unstated, so output_cost_per_second is the highest tier (720p)"},"mode":"video_generation","source":"https://fal.ai/models/veed/fabric-1.0/text","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"veed/fabric-1.0/text"},"fal_ai/wan/v2.6/image-to-image":{"mode":"image_generation","source":"https://fal.ai/models/wan/v2.6/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"wan/v2.6/image-to-image"},"fal_ai/wan/v2.6/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/wan/v2.6/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.6/image-to-video"},"fal_ai/wan/v2.6/image-to-video/flash":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/wan/v2.6/image-to-video/flash","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.6/image-to-video/flash"},"fal_ai/wan/v2.6/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/wan/v2.6/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.6/reference-to-video"},"fal_ai/wan/v2.6/text-to-image":{"mode":"image_generation","source":"https://fal.ai/models/wan/v2.6/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"wan/v2.6/text-to-image"},"fal_ai/wan/v2.6/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/wan/v2.6/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.6/text-to-video"},"fal_ai/xai/grok-imagine-image":{"mode":"image_generation","source":"https://fal.ai/models/xai/grok-imagine-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"grok-imagine-image"},"fal_ai/xai/grok-imagine-video/v1.5/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/v1.5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/v1.5/text-to-video"},"fal_ai/alibaba/happy-horse/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/alibaba/happy-horse/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"happy-horse/image-to-video"},"fal_ai/creatify/boreal":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/creatify/boreal","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"creatify/boreal"},"fal_ai/fal-ai/creatify/aurora":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/creatify/aurora","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"creatify/aurora"},"fal_ai/fal-ai/vidu/q4/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 540p $0.0315"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/q4/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/q4/image-to-video"},"fal_ai/fal-ai/vidu/q4/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; tiers without a catalog field: 540p $0.0315"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/q4/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/q4/reference-to-video"},"fal_ai/ideogram/v4.5":{"mode":"image_generation","source":"https://fal.ai/models/ideogram/v4.5","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v4.5"},"fal_ai/ideogram/v4.5/edit":{"mode":"image_generation","source":"https://fal.ai/models/ideogram/v4.5/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v4.5/edit"},"fal_ai/ideogram/v4/instant":{"metadata":{"notes":"fal bills ideogram/v4/instant at $0.0075 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/ideogram/v4/instant","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"ideogram/v4/instant"},"fal_ai/lightricks/ltx-2.5/image-to-video/fast":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.19"},"mode":"video_generation","source":"https://fal.ai/models/lightricks/ltx-2.5/image-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightricks/ltx-2.5/image-to-video/fast"},"fal_ai/lightricks/ltx-2.5/image-to-video/pro":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/lightricks/ltx-2.5/image-to-video/pro","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightricks/ltx-2.5/image-to-video/pro"},"fal_ai/lightricks/ltx-2.5/text-to-video/fast":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate; tiers without a catalog field: 1440p $0.19"},"mode":"video_generation","source":"https://fal.ai/models/lightricks/ltx-2.5/text-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightricks/ltx-2.5/text-to-video/fast"},"fal_ai/lightricks/ltx-2.5/text-to-video/pro":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/lightricks/ltx-2.5/text-to-video/pro","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"lightricks/ltx-2.5/text-to-video/pro"},"fal_ai/minimax/h3-max/recast":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/recast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/recast"},"fal_ai/minimax/h3-max/reframe-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/reframe-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/reframe-video"},"fal_ai/minimax/h3-max/relight":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/relight","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/relight"},"fal_ai/wan/v2.6/reference-to-video/flash":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/wan/v2.6/reference-to-video/flash","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.6/reference-to-video/flash"},"fal_ai/xai/grok-imagine-video/v1.5/lite/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/v1.5/lite/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/v1.5/lite/image-to-video"},"fal_ai/xai/grok-imagine-video/v1.5/lite/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/v1.5/lite/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/v1.5/lite/text-to-video"},"fal_ai/blackforestlabs/flux-3-max/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; promotional 50% off rate through 2026-11-22 23:59 UTC per the page; regular rates $0.17 (720p) and $0.29 (1080p) per second resume 2026-11-23 00:00 UTC (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3-max/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3-max/image-to-video"},"fal_ai/blackforestlabs/flux-3-max/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; promotional 50% off rate through 2026-11-22 23:59 UTC per the page; regular rates $0.17 (720p) and $0.29 (1080p) per second resume 2026-11-23 00:00 UTC (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3-max/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3-max/text-to-video"},"fal_ai/blackforestlabs/flux-3-max/keyframes-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; promotional 50% off rate through 2026-11-22 23:59 UTC per the page; regular rates $0.17 (720p) and $0.29 (1080p) per second resume 2026-11-23 00:00 UTC (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3-max/keyframes-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3-max/keyframes-to-video"},"fal_ai/heygen/heygen-video/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; regular rate; the page's 50% promotion ended October 31, 2026 (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/heygen/heygen-video/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/heygen-video/image-to-video"},"fal_ai/heygen/heygen-video/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; regular rate; the page's 50% promotion ended October 31, 2026 (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/heygen/heygen-video/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/heygen-video/reference-to-video"},"fal_ai/heygen/heygen-video/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; regular rate; the page's 50% promotion ended October 31, 2026 (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/heygen/heygen-video/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/heygen-video/text-to-video"},"fal_ai/minimax/h3-max-turbo/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; standard rate; the page's 30% promotion ran November 1 to November 15, 2026 and has ended (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max-turbo/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max-turbo/image-to-video"},"fal_ai/minimax/h3-max-turbo/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; standard rate; the page's 30% promotion ran November 1 to November 15, 2026 and has ended (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max-turbo/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max-turbo/text-to-video"},"fal_ai/minimax/h3-max/camera-controls":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 480p rate; standard rate; the page's 30% promotion ran November 1 to November 15, 2026 and has ended (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/camera-controls","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/camera-controls"},"fal_ai/minimax/h3-max/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; standard rate; the page's 30% promotion ran November 1 to November 15, 2026 and has ended (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/image-to-video"},"fal_ai/minimax/h3-max/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768p rate; standard rate; the page's 30% promotion ran November 1 to November 15, 2026 and has ended (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10)"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/text-to-video"},"replicate/alibaba/happyhorse-1.0":{"mode":"video_generation","source":"https://replicate.com/alibaba/happyhorse-1.0","provider":"replicate","base_model":"happyhorse-1.0"},"replicate/alibaba/happyhorse-1.1":{"mode":"video_generation","source":"https://replicate.com/alibaba/happyhorse-1.1","provider":"replicate","base_model":"happyhorse-1.1"},"replicate/alibaba/qwen-image-3":{"mode":"image_generation","source":"https://replicate.com/alibaba/qwen-image-3","provider":"replicate","base_model":"qwen-image-3"},"replicate/alibaba/qwen-image-3-pro":{"mode":"image_generation","source":"https://replicate.com/alibaba/qwen-image-3-pro","provider":"replicate","base_model":"qwen-image-3-pro"},"replicate/alibaba/wan-3":{"mode":"video_generation","source":"https://replicate.com/alibaba/wan-3","provider":"replicate","base_model":"wan-3"},"replicate/alibaba/wan-3-prime":{"mode":"video_generation","source":"https://replicate.com/alibaba/wan-3-prime","provider":"replicate","base_model":"wan-3-prime"},"replicate/anthropic/claude-fable-5":{"mode":"chat","source":"https://replicate.com/anthropic/claude-fable-5","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-fable-5"},"replicate/anthropic/claude-fable-5.2":{"mode":"chat","source":"https://replicate.com/anthropic/claude-fable-5.2","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-fable-5-2"},"replicate/anthropic/claude-opus-4.6":{"mode":"chat","source":"https://replicate.com/anthropic/claude-opus-4.6","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-opus-4-6"},"replicate/anthropic/claude-opus-4.7":{"mode":"chat","source":"https://replicate.com/anthropic/claude-opus-4.7","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-opus-4-7"},"replicate/anthropic/claude-opus-5.5":{"mode":"chat","source":"https://replicate.com/anthropic/claude-opus-5.5","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-opus-5-5"},"replicate/anthropic/claude-sonnet-4.6":{"mode":"chat","source":"https://replicate.com/anthropic/claude-sonnet-4.6","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-sonnet-4-6"},"replicate/anthropic/claude-sonnet-5":{"mode":"chat","source":"https://replicate.com/anthropic/claude-sonnet-5","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"claude-sonnet-5"},"replicate/black-forest-labs/flux-1.1-pro":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-1.1-pro","provider":"replicate","base_model":"flux-1.1-pro"},"replicate/black-forest-labs/flux-1.1-pro-ultra":{"mode":"image_generation","provider":"replicate","base_model":"flux-1.1-pro-ultra"},"replicate/black-forest-labs/flux-1.1-pro-ultra-finetuned":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-1.1-pro-ultra-finetuned","provider":"replicate","base_model":"flux-1.1-pro-ultra-finetuned"},"replicate/black-forest-labs/flux-2-flex":{"mode":"image_generation","provider":"replicate","base_model":"flux-2-flex"},"replicate/black-forest-labs/flux-2-klein-4b":{"mode":"image_generation","provider":"replicate","base_model":"flux-2-klein-4b"},"replicate/black-forest-labs/flux-2-klein-4b-base":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-2-klein-4b-base","provider":"replicate","base_model":"flux-2-klein-4b-base"},"replicate/black-forest-labs/flux-2-klein-4b-base-lora":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-2-klein-4b-base-lora","provider":"replicate","base_model":"flux-2-klein-4b-base-lora"},"replicate/black-forest-labs/flux-2-klein-9b":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-2-klein-9b","provider":"replicate","base_model":"flux-2-klein-9b"},"replicate/black-forest-labs/flux-2-klein-9b-base":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-2-klein-9b-base","provider":"replicate","base_model":"flux-2-klein-9b-base"},"replicate/black-forest-labs/flux-2-klein-9b-base-lora":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-2-klein-9b-base-lora","provider":"replicate","base_model":"flux-2-klein-9b-base-lora"},"replicate/black-forest-labs/flux-canny-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-canny-dev","provider":"replicate","base_model":"flux-canny-dev"},"replicate/black-forest-labs/flux-canny-pro":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-canny-pro","provider":"replicate","base_model":"flux-canny-pro"},"replicate/black-forest-labs/flux-depth-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-depth-dev","provider":"replicate","base_model":"flux-depth-dev"},"replicate/black-forest-labs/flux-depth-pro":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-depth-pro","provider":"replicate","base_model":"flux-depth-pro"},"replicate/black-forest-labs/flux-dev":{"mode":"image_generation","provider":"replicate","base_model":"flux-dev"},"replicate/black-forest-labs/flux-dev-lora":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-dev-lora","provider":"replicate","base_model":"flux-dev-lora"},"replicate/black-forest-labs/flux-fill-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-fill-dev","provider":"replicate","base_model":"flux-fill-dev"},"replicate/black-forest-labs/flux-fill-pro":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-fill-pro","provider":"replicate","base_model":"flux-fill-pro"},"replicate/black-forest-labs/flux-kontext-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-kontext-dev","provider":"replicate","base_model":"flux-kontext-dev"},"replicate/black-forest-labs/flux-kontext-dev-lora":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-kontext-dev-lora","provider":"replicate","base_model":"flux-kontext-dev-lora"},"replicate/black-forest-labs/flux-kontext-max":{"mode":"image_generation","provider":"replicate","base_model":"flux-kontext-max"},"replicate/black-forest-labs/flux-kontext-pro":{"mode":"image_generation","provider":"replicate","base_model":"flux-kontext-pro"},"replicate/black-forest-labs/flux-krea-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-krea-dev","provider":"replicate","base_model":"flux-krea-dev"},"replicate/black-forest-labs/flux-pro":{"mode":"image_generation","provider":"replicate","base_model":"flux-pro"},"replicate/black-forest-labs/flux-pro-finetuned":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-pro-finetuned","provider":"replicate","base_model":"flux-pro-finetuned"},"replicate/black-forest-labs/flux-redux-dev":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-redux-dev","provider":"replicate","base_model":"flux-redux-dev"},"replicate/black-forest-labs/flux-redux-schnell":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-redux-schnell","provider":"replicate","base_model":"flux-redux-schnell"},"replicate/black-forest-labs/flux-schnell":{"mode":"image_generation","provider":"replicate","base_model":"flux-schnell"},"replicate/black-forest-labs/flux-schnell-lora":{"mode":"image_generation","source":"https://replicate.com/black-forest-labs/flux-schnell-lora","provider":"replicate","base_model":"flux-schnell-lora"},"replicate/bria/eraser":{"mode":"image_generation","source":"https://replicate.com/bria/eraser","provider":"replicate","base_model":"bria/eraser"},"replicate/bria/expand-image":{"mode":"image_generation","source":"https://replicate.com/bria/expand-image","provider":"replicate","base_model":"bria/expand-image"},"replicate/bria/fibo":{"mode":"image_generation","source":"https://replicate.com/bria/fibo","provider":"replicate","base_model":"bria/fibo"},"replicate/bria/fibo-edit":{"mode":"image_generation","source":"https://replicate.com/bria/fibo-edit","provider":"replicate","base_model":"bria/fibo-edit"},"replicate/bria/generate-background":{"mode":"image_generation","source":"https://replicate.com/bria/generate-background","provider":"replicate","base_model":"bria/generate-background"},"replicate/bria/genfill":{"mode":"image_generation","source":"https://replicate.com/bria/genfill","provider":"replicate","base_model":"bria/genfill"},"replicate/bria/image-3.2":{"mode":"image_generation","source":"https://replicate.com/bria/image-3.2","provider":"replicate","base_model":"bria/image-3.2"},"replicate/bria/increase-resolution":{"mode":"image_generation","source":"https://replicate.com/bria/increase-resolution","provider":"replicate","base_model":"bria/increase-resolution"},"replicate/bria/product-cutout":{"mode":"image_generation","source":"https://replicate.com/bria/product-cutout","provider":"replicate","base_model":"bria/product-cutout"},"replicate/bria/product-packshot":{"mode":"image_generation","source":"https://replicate.com/bria/product-packshot","provider":"replicate","base_model":"bria/product-packshot"},"replicate/bria/product-shadow":{"mode":"image_generation","source":"https://replicate.com/bria/product-shadow","provider":"replicate","base_model":"bria/product-shadow"},"replicate/bria/remove-background":{"mode":"image_generation","source":"https://replicate.com/bria/remove-background","provider":"replicate","base_model":"bria/remove-background"},"replicate/bria/video-erase-object":{"mode":"video_generation","source":"https://replicate.com/bria/video-erase-object","provider":"replicate","base_model":"bria/video-erase-object"},"replicate/bria/video-increase-resolution":{"mode":"video_generation","source":"https://replicate.com/bria/video-increase-resolution","provider":"replicate","base_model":"bria/video-increase-resolution"},"replicate/bria/video-remove-background":{"mode":"video_generation","source":"https://replicate.com/bria/video-remove-background","provider":"replicate","base_model":"bria/video-remove-background"},"replicate/bytedance/dreamactor-m2.0":{"mode":"video_generation","source":"https://replicate.com/bytedance/dreamactor-m2.0","provider":"replicate","base_model":"dreamactor-m2.0"},"replicate/bytedance/dreamina-3.1":{"mode":"image_generation","source":"https://replicate.com/bytedance/dreamina-3.1","provider":"replicate","base_model":"dreamina-3.1"},"replicate/bytedance/omni-human":{"mode":"video_generation","source":"https://replicate.com/bytedance/omni-human","provider":"replicate","base_model":"omni-human"},"replicate/bytedance/omni-human-1.5":{"mode":"video_generation","source":"https://replicate.com/bytedance/omni-human-1.5","provider":"replicate","base_model":"omni-human-1.5"},"replicate/bytedance/seedance-1-lite":{"mode":"video_generation","provider":"replicate","base_model":"seedance-1-lite"},"replicate/bytedance/seedance-1-pro":{"mode":"video_generation","provider":"replicate","base_model":"seedance-1-pro"},"replicate/bytedance/seedance-1-pro-fast":{"mode":"video_generation","source":"https://replicate.com/bytedance/seedance-1-pro-fast","provider":"replicate","base_model":"seedance-1-pro-fast"},"replicate/bytedance/seedream-3":{"mode":"image_generation","source":"https://replicate.com/bytedance/seedream-3","provider":"replicate","base_model":"seedream-3"},"replicate/bytedance/seedream-4":{"mode":"image_generation","provider":"replicate","base_model":"seedream-4"},"replicate/bytedance/seedream-4.5":{"mode":"image_generation","source":"https://replicate.com/bytedance/seedream-4.5","provider":"replicate","base_model":"seedream-4.5"},"replicate/bytedance/seedream-5-lite":{"mode":"image_generation","provider":"replicate","base_model":"seedream-5-lite"},"replicate/datalab-to/ocr":{"mode":"ocr","source":"https://replicate.com/datalab-to/ocr","provider":"replicate","base_model":"datalab-to/ocr"},"replicate/decart/lucy-edit-2":{"mode":"video_generation","source":"https://replicate.com/decart/lucy-edit-2","provider":"replicate","base_model":"decart/lucy-edit-2"},"replicate/easel/ai-avatars":{"mode":"image_generation","source":"https://replicate.com/easel/ai-avatars","provider":"replicate","base_model":"easel/ai-avatars"},"replicate/elevenlabs/dubbing":{"mode":"audio_speech","source":"https://replicate.com/elevenlabs/dubbing","provider":"replicate","base_model":"dubbing"},"replicate/elevenlabs/flash-v2.5":{"mode":"audio_speech","source":"https://replicate.com/elevenlabs/flash-v2.5","provider":"replicate","base_model":"flash"},"replicate/elevenlabs/turbo-v2.5":{"mode":"audio_speech","source":"https://replicate.com/elevenlabs/turbo-v2.5","provider":"replicate","base_model":"turbo"},"replicate/elevenlabs/v2-multilingual":{"mode":"audio_speech","source":"https://replicate.com/elevenlabs/v2-multilingual","provider":"replicate","base_model":"v2-multilingual"},"replicate/elevenlabs/v3":{"mode":"audio_speech","source":"https://replicate.com/elevenlabs/v3","provider":"replicate","base_model":"v3"},"replicate/flux-kontext-apps/cartoonify":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/cartoonify","provider":"replicate","base_model":"flux-kontext-apps/cartoonify"},"replicate/flux-kontext-apps/change-haircut":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/change-haircut","provider":"replicate","base_model":"flux-kontext-apps/change-haircut"},"replicate/flux-kontext-apps/depth-of-field":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/depth-of-field","provider":"replicate","base_model":"flux-kontext-apps/depth-of-field"},"replicate/flux-kontext-apps/face-to-many-kontext":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/face-to-many-kontext","provider":"replicate","base_model":"flux-kontext-apps/face-to-many-kontext"},"replicate/flux-kontext-apps/filters":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/filters","provider":"replicate","base_model":"flux-kontext-apps/filters"},"replicate/flux-kontext-apps/iconic-locations":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/iconic-locations","provider":"replicate","base_model":"flux-kontext-apps/iconic-locations"},"replicate/flux-kontext-apps/impossible-scenarios":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/impossible-scenarios","provider":"replicate","base_model":"flux-kontext-apps/impossible-scenarios"},"replicate/flux-kontext-apps/multi-image-kontext-max":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/multi-image-kontext-max","provider":"replicate","base_model":"flux-kontext-apps/multi-image-kontext-max"},"replicate/flux-kontext-apps/multi-image-kontext-pro":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/multi-image-kontext-pro","provider":"replicate","base_model":"flux-kontext-apps/multi-image-kontext-pro"},"replicate/flux-kontext-apps/multi-image-list":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/multi-image-list","provider":"replicate","base_model":"flux-kontext-apps/multi-image-list"},"replicate/flux-kontext-apps/portrait-series":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/portrait-series","provider":"replicate","base_model":"flux-kontext-apps/portrait-series"},"replicate/flux-kontext-apps/professional-headshot":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/professional-headshot","provider":"replicate","base_model":"flux-kontext-apps/professional-headshot"},"replicate/flux-kontext-apps/restore-image":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/restore-image","provider":"replicate","base_model":"flux-kontext-apps/restore-image"},"replicate/flux-kontext-apps/restyle-video-frame":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/restyle-video-frame","provider":"replicate","base_model":"flux-kontext-apps/restyle-video-frame"},"replicate/flux-kontext-apps/text-removal":{"mode":"image_generation","source":"https://replicate.com/flux-kontext-apps/text-removal","provider":"replicate","base_model":"flux-kontext-apps/text-removal"},"replicate/google/gemini-2.5-flash-image":{"mode":"image_generation","source":"https://replicate.com/google/gemini-2.5-flash-image","provider":"replicate","base_model":"gemini-2.5-flash-image"},"replicate/google/gemini-3-flash":{"mode":"chat","source":"https://replicate.com/google/gemini-3-flash","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gemini-3-flash"},"replicate/google/gemini-3.1-flash-tts":{"mode":"audio_speech","source":"https://replicate.com/google/gemini-3.1-flash-tts","provider":"replicate","base_model":"gemini-3.1-flash-tts"},"replicate/google/gemini-3.1-pro":{"mode":"chat","source":"https://replicate.com/google/gemini-3.1-pro","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gemini-3.1-pro"},"replicate/google/gemini-3.5-flash":{"mode":"chat","source":"https://replicate.com/google/gemini-3.5-flash","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gemini-3.5-flash"},"replicate/google/gemini-3.8-flash-lite-tts":{"mode":"audio_speech","source":"https://replicate.com/google/gemini-3.8-flash-lite-tts","provider":"replicate","base_model":"gemini-3.8-flash-lite-tts"},"replicate/google/gemini-3.8-flash-tts":{"mode":"audio_speech","source":"https://replicate.com/google/gemini-3.8-flash-tts","provider":"replicate","base_model":"gemini-3.8-flash-tts"},"replicate/google/nano-banana":{"mode":"image_generation","provider":"replicate","base_model":"nano-banana"},"replicate/google/nano-banana-2":{"mode":"image_generation","source":"https://replicate.com/google/nano-banana-2","provider":"replicate","base_model":"nano-banana-2"},"replicate/google/nano-banana-2-lite":{"mode":"image_generation","source":"https://replicate.com/google/nano-banana-2-lite","provider":"replicate","base_model":"nano-banana-2-lite"},"replicate/google/nano-banana-2.1":{"mode":"image_generation","source":"https://replicate.com/google/nano-banana-2.1","provider":"replicate","base_model":"nano-banana-2.1"},"replicate/google/nano-banana-latest":{"alias_of":"replicate/google/nano-banana-2.1","base_model":"nano-banana"},"replicate/heygen/avatar-iv":{"mode":"video_generation","source":"https://replicate.com/heygen/avatar-iv","provider":"replicate","base_model":"heygen/avatar-iv"},"replicate/heygen/avatar-v":{"mode":"video_generation","source":"https://replicate.com/heygen/avatar-v","provider":"replicate","base_model":"heygen/avatar-v"},"replicate/heygen/lipsync-precision":{"mode":"video_generation","source":"https://replicate.com/heygen/lipsync-precision","provider":"replicate","base_model":"heygen/lipsync-precision"},"replicate/heygen/lipsync-speed":{"mode":"video_generation","source":"https://replicate.com/heygen/lipsync-speed","provider":"replicate","base_model":"heygen/lipsync-speed"},"replicate/heygen/video-agent":{"mode":"video_generation","source":"https://replicate.com/heygen/video-agent","provider":"replicate","base_model":"heygen/video-agent"},"replicate/ibm-granite/granite-20b-code-instruct-8k":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-20b-code-instruct-8k","supports_system_messages":true,"provider":"replicate","base_model":"granite-20b-code-instruct-8k"},"replicate/ibm-granite/granite-3.0-2b-instruct":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-3.0-2b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"granite-3.0-2b-instruct"},"replicate/ibm-granite/granite-3.0-8b-instruct":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-3.0-8b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"granite-3.0-8b-instruct"},"replicate/ibm-granite/granite-3.1-2b-instruct":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-3.1-2b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"granite-3.1-2b-instruct"},"replicate/ibm-granite/granite-3.1-8b-instruct":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-3.1-8b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"granite-3.1-8b-instruct"},"replicate/ibm-granite/granite-3.2-8b-instruct":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-3.2-8b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"granite-3.2-8b-instruct"},"replicate/ibm-granite/granite-4.0-h-small":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-4.0-h-small","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","base_model":"granite-4.0-h-small"},"replicate/ibm-granite/granite-4.1-8b":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-4.1-8b","supports_function_calling":true,"supports_system_messages":true,"provider":"replicate","base_model":"granite-4.1-8b"},"replicate/ibm-granite/granite-4.2-8b":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-4.2-8b","supports_function_calling":true,"supports_reasoning":true,"supports_system_messages":true,"provider":"replicate","base_model":"granite-4.2-8b"},"replicate/ibm-granite/granite-8b-code-instruct-128k":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-8b-code-instruct-128k","supports_system_messages":true,"provider":"replicate","base_model":"granite-8b-code-instruct-128k"},"replicate/ibm-granite/granite-speech-3.3-8b":{"mode":"audio_transcription","source":"https://replicate.com/ibm-granite/granite-speech-3.3-8b","provider":"replicate","base_model":"granite-speech-3.3-8b"},"replicate/ibm-granite/granite-speech-4.1-2b":{"mode":"audio_transcription","source":"https://replicate.com/ibm-granite/granite-speech-4.1-2b","provider":"replicate","base_model":"granite-speech-4.1-2b"},"replicate/ibm-granite/granite-vision-3.3-2b":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-vision-3.3-2b","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"granite-vision-3.3-2b"},"replicate/ibm-granite/granite-vision-4.1-4b":{"mode":"chat","source":"https://replicate.com/ibm-granite/granite-vision-4.1-4b","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"granite-vision-4.1-4b"},"replicate/ideogram-ai/ideogram-v2":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v2","provider":"replicate","base_model":"ideogram-ai/ideogram-v2"},"replicate/ideogram-ai/ideogram-v2-turbo":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v2-turbo","provider":"replicate","base_model":"ideogram-ai/ideogram-v2-turbo"},"replicate/ideogram-ai/ideogram-v2a":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v2a","provider":"replicate","base_model":"ideogram-ai/ideogram-v2a"},"replicate/ideogram-ai/ideogram-v2a-turbo":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v2a-turbo","provider":"replicate","base_model":"ideogram-ai/ideogram-v2a-turbo"},"replicate/ideogram-ai/ideogram-v3-balanced":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v3-balanced","provider":"replicate","base_model":"ideogram-v3-balanced"},"replicate/ideogram-ai/ideogram-v3-quality":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v3-quality","provider":"replicate","base_model":"ideogram-ai/ideogram-v3-quality"},"replicate/ideogram-ai/ideogram-v3-turbo":{"mode":"image_generation","provider":"replicate","base_model":"ideogram-ai/ideogram-v3-turbo"},"replicate/ideogram-ai/ideogram-v4-balanced":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v4-balanced","provider":"replicate","base_model":"ideogram-ai/ideogram-v4-balanced"},"replicate/ideogram-ai/ideogram-v4-quality":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/ideogram-v4-quality","provider":"replicate","base_model":"ideogram-ai/ideogram-v4-quality"},"replicate/ideogram-ai/layerize":{"mode":"image_generation","source":"https://replicate.com/ideogram-ai/layerize","provider":"replicate","base_model":"ideogram-ai/layerize"},"replicate/intelligent-utilities/html-to-image":{"mode":"image_generation","source":"https://replicate.com/intelligent-utilities/html-to-image","provider":"replicate","base_model":"intelligent-utilities/html-to-image"},"replicate/inworld/realtime-tts-1.5-max":{"mode":"audio_speech","source":"https://replicate.com/inworld/realtime-tts-1.5-max","provider":"replicate","base_model":"inworld/realtime-tts-1.5-max"},"replicate/inworld/realtime-tts-1.5-mini":{"mode":"audio_speech","source":"https://replicate.com/inworld/realtime-tts-1.5-mini","provider":"replicate","base_model":"inworld/realtime-tts-1.5-mini"},"replicate/inworld/realtime-tts-2":{"mode":"audio_speech","source":"https://replicate.com/inworld/realtime-tts-2","provider":"replicate","base_model":"inworld/realtime-tts-2"},"replicate/kwaivgi/kling-lip-sync":{"mode":"video_generation","source":"https://replicate.com/kwaivgi/kling-lip-sync","provider":"replicate","base_model":"kwaivgi/kling-lip-sync"},"replicate/kwaivgi/kling-v2.1-master":{"mode":"video_generation","provider":"replicate","base_model":"kwaivgi/kling-v2.1-master"},"replicate/kwaivgi/kling-v2.5-turbo-pro":{"mode":"video_generation","source":"https://replicate.com/kwaivgi/kling-v2.5-turbo-pro","provider":"replicate","base_model":"kwaivgi/kling-v2.5-turbo-pro"},"replicate/lightricks/audio-to-video":{"mode":"video_generation","source":"https://replicate.com/lightricks/audio-to-video","provider":"replicate","base_model":"lightricks/audio-to-video"},"replicate/lightricks/ltx-2-distilled":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2-distilled","provider":"replicate","base_model":"lightricks/ltx-2-distilled"},"replicate/lightricks/ltx-2-fast":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2-fast","provider":"replicate","base_model":"lightricks/ltx-2-fast"},"replicate/lightricks/ltx-2-pro":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2-pro","provider":"replicate","base_model":"lightricks/ltx-2-pro"},"replicate/lightricks/ltx-2-retake":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2-retake","provider":"replicate","base_model":"lightricks/ltx-2-retake"},"replicate/lightricks/ltx-2.3-fast":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2.3-fast","provider":"replicate","base_model":"lightricks/ltx-2.3-fast"},"replicate/lightricks/ltx-2.3-pro":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2.3-pro","provider":"replicate","base_model":"lightricks/ltx-2.3-pro"},"replicate/lightricks/ltx-2.5-fast":{"mode":"video_generation","source":"https://replicate.com/lightricks/ltx-2.5-fast","provider":"replicate","base_model":"lightricks/ltx-2.5-fast"},"replicate/luma/photon":{"mode":"image_generation","source":"https://replicate.com/luma/photon","provider":"replicate","base_model":"luma/photon"},"replicate/luma/photon-flash":{"mode":"image_generation","source":"https://replicate.com/luma/photon-flash","provider":"replicate","base_model":"luma/photon-flash"},"replicate/luma/reframe-video":{"mode":"video_generation","source":"https://replicate.com/luma/reframe-video","provider":"replicate","base_model":"luma/reframe-video"},"replicate/meta/llama-4-maverick-instruct":{"mode":"chat","source":"https://replicate.com/meta/llama-4-maverick-instruct","supports_system_messages":true,"provider":"replicate","base_model":"llama-4-maverick-instruct"},"replicate/meta/llama-4-scout-instruct":{"mode":"chat","source":"https://replicate.com/meta/llama-4-scout-instruct","supports_system_messages":true,"provider":"replicate","base_model":"llama-4-scout-instruct"},"replicate/meta/llama-guard-4-12b":{"mode":"chat","source":"https://replicate.com/meta/llama-guard-4-12b","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"llama-guard-4-12b"},"replicate/meta/meta-llama-3-70b":{"mode":"chat","source":"https://replicate.com/meta/meta-llama-3-70b","provider":"replicate","base_model":"llama-3-70b"},"replicate/meta/meta-llama-3-70b-instruct":{"mode":"chat","source":"https://replicate.com/meta/meta-llama-3-70b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"llama-3-70b-instruct"},"replicate/meta/meta-llama-3-8b":{"mode":"chat","source":"https://replicate.com/meta/meta-llama-3-8b","provider":"replicate","base_model":"llama-3-8b"},"replicate/meta/meta-llama-3-8b-instruct":{"mode":"chat","source":"https://replicate.com/meta/meta-llama-3-8b-instruct","supports_system_messages":true,"provider":"replicate","base_model":"llama-3-8b-instruct"},"replicate/minimax/h3":{"mode":"video_generation","source":"https://replicate.com/minimax/h3","provider":"replicate","base_model":"h3"},"replicate/minimax/h3-highspeed":{"mode":"video_generation","source":"https://replicate.com/minimax/h3-highspeed","provider":"replicate","base_model":"h3-highspeed"},"replicate/minimax/h3-lite":{"mode":"video_generation","source":"https://replicate.com/minimax/h3-lite","provider":"replicate","base_model":"h3-lite"},"replicate/minimax/image-01":{"mode":"image_generation","source":"https://replicate.com/minimax/image-01","provider":"replicate","base_model":"image-01"},"replicate/minimax/speech-02-hd":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-02-hd","provider":"replicate","base_model":"speech-02-hd"},"replicate/minimax/speech-02-turbo":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-02-turbo","provider":"replicate","base_model":"speech-02-turbo"},"replicate/minimax/speech-2.6-hd":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-2.6-hd","provider":"replicate","base_model":"speech-2.6-hd"},"replicate/minimax/speech-2.6-turbo":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-2.6-turbo","provider":"replicate","base_model":"speech-2.6-turbo"},"replicate/minimax/speech-2.8-hd":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-2.8-hd","provider":"replicate","base_model":"speech-2.8-hd"},"replicate/minimax/speech-2.8-turbo":{"mode":"audio_speech","source":"https://replicate.com/minimax/speech-2.8-turbo","provider":"replicate","base_model":"speech-2.8-turbo"},"replicate/moonshotai/kimi-k2-thinking":{"mode":"chat","source":"https://replicate.com/moonshotai/kimi-k2-thinking","provider":"replicate","base_model":"kimi-k2-thinking"},"replicate/moonshotai/kimi-k2.5":{"mode":"chat","source":"https://replicate.com/moonshotai/kimi-k2.5","supports_vision":true,"provider":"replicate","base_model":"kimi-k2.5"},"replicate/moonshotai/kimi-k2.6":{"mode":"chat","source":"https://replicate.com/moonshotai/kimi-k2.6","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"kimi-k2.6"},"replicate/nightmareai/real-esrgan":{"mode":"image_generation","source":"https://replicate.com/nightmareai/real-esrgan","provider":"replicate","base_model":"nightmareai/real-esrgan"},"replicate/openai/gpt-4o-mini-transcribe":{"mode":"audio_transcription","source":"https://replicate.com/openai/gpt-4o-mini-transcribe","provider":"replicate","base_model":"gpt-4o-mini-transcribe"},"replicate/openai/gpt-4o-transcribe":{"mode":"audio_transcription","source":"https://replicate.com/openai/gpt-4o-transcribe","provider":"replicate","base_model":"gpt-4o-transcribe"},"replicate/openai/gpt-5-pro":{"mode":"chat","source":"https://replicate.com/openai/gpt-5-pro","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5-pro"},"replicate/openai/gpt-5.1":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.1","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.1"},"replicate/openai/gpt-5.2":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.2","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.2"},"replicate/openai/gpt-5.4":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.4","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.4"},"replicate/openai/gpt-5.6-luna":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.6-luna","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.6-luna"},"replicate/openai/gpt-5.6-sol":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.6-sol","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.6-sol"},"replicate/openai/gpt-5.6-terra":{"mode":"chat","source":"https://replicate.com/openai/gpt-5.6-terra","supports_reasoning":true,"supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"gpt-5.6-terra"},"replicate/perceptron-ai-inc/isaac-0.1":{"mode":"chat","source":"https://replicate.com/perceptron-ai-inc/isaac-0.1","supports_vision":true,"provider":"replicate","base_model":"perceptron-ai-inc/isaac-0.1"},"replicate/philz1337x/clarity-pro-upscaler":{"mode":"image_generation","source":"https://replicate.com/philz1337x/clarity-pro-upscaler","provider":"replicate","base_model":"philz1337x/clarity-pro-upscaler"},"replicate/pixverse/lipsync":{"mode":"video_generation","source":"https://replicate.com/pixverse/lipsync","provider":"replicate","base_model":"pixverse/lipsync"},"replicate/playht/play-dialog":{"mode":"audio_speech","source":"https://replicate.com/playht/play-dialog","provider":"replicate","base_model":"playht/play-dialog"},"replicate/prunaai/flux-fast":{"mode":"image_generation","provider":"replicate","base_model":"prunaai/flux-fast"},"replicate/prunaai/flux-kontext-fast":{"mode":"image_generation","source":"https://replicate.com/prunaai/flux-kontext-fast","provider":"replicate","base_model":"prunaai/flux-kontext-fast"},"replicate/prunaai/hidream-l1-fast":{"mode":"image_generation","provider":"replicate","base_model":"prunaai/hidream-l1-fast"},"replicate/prunaai/p-image":{"mode":"image_generation","provider":"replicate","base_model":"prunaai/p-image"},"replicate/prunaai/p-image-edit":{"mode":"image_generation","source":"https://replicate.com/prunaai/p-image-edit","provider":"replicate","base_model":"prunaai/p-image-edit"},"replicate/prunaai/p-image-edit-lora":{"mode":"image_generation","source":"https://replicate.com/prunaai/p-image-edit-lora","provider":"replicate","base_model":"prunaai/p-image-edit-lora"},"replicate/prunaai/p-image-lora":{"mode":"image_generation","source":"https://replicate.com/prunaai/p-image-lora","provider":"replicate","base_model":"prunaai/p-image-lora"},"replicate/prunaai/p-video-2-pro-avatar":{"mode":"video_generation","source":"https://replicate.com/prunaai/p-video-2-pro-avatar","provider":"replicate","base_model":"prunaai/p-video-2-pro-avatar"},"replicate/prunaai/p-video-2-pro-replace":{"mode":"video_generation","source":"https://replicate.com/prunaai/p-video-2-pro-replace","provider":"replicate","base_model":"prunaai/p-video-2-pro-replace"},"replicate/prunaai/p-video-animate":{"mode":"video_generation","source":"https://replicate.com/prunaai/p-video-animate","provider":"replicate","base_model":"prunaai/p-video-animate"},"replicate/prunaai/p-video-avatar":{"mode":"video_generation","source":"https://replicate.com/prunaai/p-video-avatar","provider":"replicate","base_model":"prunaai/p-video-avatar"},"replicate/prunaai/p-video-replace":{"mode":"video_generation","source":"https://replicate.com/prunaai/p-video-replace","provider":"replicate","base_model":"prunaai/p-video-replace"},"replicate/prunaai/wan-2.2-image":{"mode":"image_generation","source":"https://replicate.com/prunaai/wan-2.2-image","provider":"replicate","base_model":"prunaai/wan-2.2-image"},"replicate/quiverai/arrow-1.1":{"mode":"image_generation","source":"https://replicate.com/quiverai/arrow-1.1","provider":"replicate","base_model":"quiverai/arrow-1.1"},"replicate/quiverai/arrow-1.1-max":{"mode":"image_generation","source":"https://replicate.com/quiverai/arrow-1.1-max","provider":"replicate","base_model":"quiverai/arrow-1.1-max"},"replicate/quiverai/arrow-1.1-max-vectorize":{"mode":"image_generation","source":"https://replicate.com/quiverai/arrow-1.1-max-vectorize","provider":"replicate","base_model":"quiverai/arrow-1.1-max-vectorize"},"replicate/quiverai/arrow-1.1-vectorize":{"mode":"image_generation","source":"https://replicate.com/quiverai/arrow-1.1-vectorize","provider":"replicate","base_model":"quiverai/arrow-1.1-vectorize"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-fusion":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-fusion","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-fusion"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-next-scene":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-next-scene","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-next-scene"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-photo-to-anime":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-photo-to-anime","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-photo-to-anime"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-relight":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-relight","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-relight"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-skin":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-skin","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-skin"},"replicate/qwen-edit-apps/qwen-image-edit-plus-lora-upscale":{"mode":"image_generation","source":"https://replicate.com/qwen-edit-apps/qwen-image-edit-plus-lora-upscale","provider":"replicate","base_model":"qwen-edit-apps/qwen-image-edit-plus-lora-upscale"},"replicate/qwen/qwen-edit-multiangle":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-edit-multiangle","provider":"replicate","base_model":"qwen-edit-multiangle"},"replicate/qwen/qwen-image":{"mode":"image_generation","provider":"replicate","base_model":"qwen-image"},"replicate/qwen/qwen-image-2":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-2","provider":"replicate","base_model":"qwen-image-2"},"replicate/qwen/qwen-image-2-pro":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-2-pro","provider":"replicate","base_model":"qwen-image-2-pro"},"replicate/qwen/qwen-image-2512":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-2512","provider":"replicate","base_model":"qwen-image"},"replicate/qwen/qwen-image-edit":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-edit","provider":"replicate","base_model":"qwen-image-edit"},"replicate/qwen/qwen-image-edit-2511":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-edit-2511","provider":"replicate","base_model":"qwen-image-edit-2511"},"replicate/qwen/qwen-image-edit-plus":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-edit-plus","provider":"replicate","base_model":"qwen-image-edit-plus"},"replicate/qwen/qwen-image-edit-plus-lora":{"mode":"image_generation","source":"https://replicate.com/qwen/qwen-image-edit-plus-lora","provider":"replicate","base_model":"qwen-image-edit-plus-lora"},"replicate/qwen/qwen3-7-plus":{"mode":"chat","source":"https://replicate.com/qwen/qwen3-7-plus","supports_system_messages":true,"supports_vision":true,"provider":"replicate","base_model":"qwen3-7-plus"},"replicate/qwen/qwen3-tts":{"mode":"audio_speech","source":"https://replicate.com/qwen/qwen3-tts","provider":"replicate","base_model":"qwen3-tts"},"replicate/recraft-ai/recraft-20b":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-20b","provider":"replicate","base_model":"recraft-ai/recraft-20b"},"replicate/recraft-ai/recraft-20b-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-20b-svg","provider":"replicate","base_model":"recraft-ai/recraft-20b-svg"},"replicate/recraft-ai/recraft-creative-upscale":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-creative-upscale","provider":"replicate","base_model":"recraft-ai/recraft-creative-upscale"},"replicate/recraft-ai/recraft-crisp-upscale":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-crisp-upscale","provider":"replicate","base_model":"recraft-ai/recraft-crisp-upscale"},"replicate/recraft-ai/recraft-remove-background":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-remove-background","provider":"replicate","base_model":"recraft-ai/recraft-remove-background"},"replicate/recraft-ai/recraft-v3":{"mode":"image_generation","provider":"replicate","base_model":"recraft-ai/recraft-v3"},"replicate/recraft-ai/recraft-v3-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v3-svg","provider":"replicate","base_model":"recraft-ai/recraft-v3-svg"},"replicate/recraft-ai/recraft-v4":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4","provider":"replicate","base_model":"recraft-ai/recraft-v4"},"replicate/recraft-ai/recraft-v4-pro":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4-pro","provider":"replicate","base_model":"recraft-ai/recraft-v4-pro"},"replicate/recraft-ai/recraft-v4-pro-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4-pro-svg","provider":"replicate","base_model":"recraft-ai/recraft-v4-pro-svg"},"replicate/recraft-ai/recraft-v4-styles-pro":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4-styles-pro","provider":"replicate","base_model":"recraft-ai/recraft-v4-styles-pro"},"replicate/recraft-ai/recraft-v4-styles-pro-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4-styles-pro-svg","provider":"replicate","base_model":"recraft-ai/recraft-v4-styles-pro-svg"},"replicate/recraft-ai/recraft-v4-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4-svg","provider":"replicate","base_model":"recraft-ai/recraft-v4-svg"},"replicate/recraft-ai/recraft-v4.1":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1","provider":"replicate","base_model":"recraft-v4.1"},"replicate/recraft-ai/recraft-v4.1-pro":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1-pro","provider":"replicate","base_model":"recraft-ai/recraft-v4.1-pro"},"replicate/recraft-ai/recraft-v4.1-pro-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1-pro-svg","provider":"replicate","base_model":"recraft-ai/recraft-v4.1-pro-svg"},"replicate/recraft-ai/recraft-v4.1-svg":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1-svg","provider":"replicate","base_model":"recraft-ai/recraft-v4.1-svg"},"replicate/recraft-ai/recraft-v4.1-utility":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1-utility","provider":"replicate","base_model":"recraft-v4.1-utility"},"replicate/recraft-ai/recraft-v4.1-utility-pro":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-v4.1-utility-pro","provider":"replicate","base_model":"recraft-ai/recraft-v4.1-utility-pro"},"replicate/recraft-ai/recraft-vectorize":{"mode":"image_generation","source":"https://replicate.com/recraft-ai/recraft-vectorize","provider":"replicate","base_model":"recraft-vectorize"},"replicate/resemble-ai/chatterbox":{"mode":"audio_speech","source":"https://replicate.com/resemble-ai/chatterbox","provider":"replicate","base_model":"resemble-ai/chatterbox"},"replicate/resemble-ai/chatterbox-pro":{"mode":"audio_speech","source":"https://replicate.com/resemble-ai/chatterbox-pro","provider":"replicate","base_model":"resemble-ai/chatterbox-pro"},"replicate/resemble-ai/chatterbox-turbo":{"mode":"audio_speech","source":"https://replicate.com/resemble-ai/chatterbox-turbo","provider":"replicate","base_model":"resemble-ai/chatterbox-turbo"},"replicate/retro-diffusion/rd-animation":{"mode":"image_generation","source":"https://replicate.com/retro-diffusion/rd-animation","provider":"replicate","base_model":"retro-diffusion/rd-animation"},"replicate/runwayml/aleph-2":{"mode":"video_generation","source":"https://replicate.com/runwayml/aleph-2","provider":"replicate","base_model":"aleph-2"},"replicate/runwayml/gen-4.5":{"mode":"video_generation","source":"https://replicate.com/runwayml/gen-4.5","provider":"replicate","base_model":"gen-4.5"},"replicate/runwayml/gen4-image-turbo":{"mode":"image_generation","source":"https://replicate.com/runwayml/gen4-image-turbo","provider":"replicate","base_model":"gen4-image-turbo"},"replicate/runwayml/gen4-turbo":{"mode":"video_generation","provider":"replicate","base_model":"gen4-turbo"},"replicate/snowflake/snowflake-arctic-instruct":{"mode":"chat","source":"https://replicate.com/snowflake/snowflake-arctic-instruct","supports_system_messages":true,"provider":"replicate","base_model":"snowflake-arctic-instruct"},"replicate/stability-ai/stable-diffusion-3":{"mode":"image_generation","source":"https://replicate.com/stability-ai/stable-diffusion-3","provider":"replicate","base_model":"stability-ai/stable-diffusion-3"},"replicate/stability-ai/stable-diffusion-3.5-large":{"mode":"image_generation","provider":"replicate","base_model":"stability-ai/stable-diffusion-3.5-large"},"replicate/stability-ai/stable-diffusion-3.5-large-turbo":{"mode":"image_generation","source":"https://replicate.com/stability-ai/stable-diffusion-3.5-large-turbo","provider":"replicate","base_model":"stability-ai/stable-diffusion-3.5-large-turbo"},"replicate/stability-ai/stable-diffusion-3.5-medium":{"mode":"image_generation","provider":"replicate","base_model":"stability-ai/stable-diffusion-3.5-medium"},"replicate/sync/lipsync-2":{"mode":"video_generation","source":"https://replicate.com/sync/lipsync-2","provider":"replicate","base_model":"sync/lipsync-2"},"replicate/sync/lipsync-2-pro":{"mode":"video_generation","source":"https://replicate.com/sync/lipsync-2-pro","provider":"replicate","base_model":"sync/lipsync-2-pro"},"replicate/sync/react-1":{"mode":"video_generation","source":"https://replicate.com/sync/react-1","provider":"replicate","base_model":"sync/react-1"},"replicate/tencent/hunyuan-image-2.1":{"mode":"image_generation","source":"https://replicate.com/tencent/hunyuan-image-2.1","provider":"replicate","base_model":"hunyuan-image-2.1"},"replicate/tencent/hunyuan-image-3":{"mode":"image_generation","source":"https://replicate.com/tencent/hunyuan-image-3","provider":"replicate","base_model":"hunyuan-image-3"},"replicate/veed/fabric-1.0":{"mode":"video_generation","source":"https://replicate.com/veed/fabric-1.0","provider":"replicate","base_model":"veed/fabric-1.0"},"replicate/wan-video/wan-2.2-s2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.2-s2v","provider":"replicate","base_model":"wan-video/wan-2.2-s2v"},"replicate/wan-video/wan-2.5-i2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.5-i2v","provider":"replicate","base_model":"wan-video/wan-2.5-i2v"},"replicate/wan-video/wan-2.5-i2v-fast":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.5-i2v-fast","provider":"replicate","base_model":"wan-video/wan-2.5-i2v-fast"},"replicate/wan-video/wan-2.5-t2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.5-t2v","provider":"replicate","base_model":"wan-video/wan-2.5-t2v"},"replicate/wan-video/wan-2.5-t2v-fast":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.5-t2v-fast","provider":"replicate","base_model":"wan-video/wan-2.5-t2v-fast"},"replicate/wan-video/wan-2.6-i2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.6-i2v","provider":"replicate","base_model":"wan-video/wan-2.6-i2v"},"replicate/wan-video/wan-2.6-t2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.6-t2v","provider":"replicate","base_model":"wan-2.6-t2v"},"replicate/wan-video/wan-2.7-i2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.7-i2v","provider":"replicate","base_model":"wan-video/wan-2.7-i2v"},"replicate/wan-video/wan-2.7-image":{"mode":"image_generation","source":"https://replicate.com/wan-video/wan-2.7-image","provider":"replicate","base_model":"wan-video/wan-2.7-image"},"replicate/wan-video/wan-2.7-image-pro":{"mode":"image_generation","source":"https://replicate.com/wan-video/wan-2.7-image-pro","provider":"replicate","base_model":"wan-video/wan-2.7-image-pro"},"replicate/wan-video/wan-2.7-r2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.7-r2v","provider":"replicate","base_model":"wan-video/wan-2.7-r2v"},"replicate/wan-video/wan-2.7-t2v":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.7-t2v","provider":"replicate","base_model":"wan-2.7-t2v"},"replicate/wan-video/wan-2.7-videoedit":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan-2.7-videoedit","provider":"replicate","base_model":"wan-video/wan-2.7-videoedit"},"replicate/wan-video/wan2.6-i2v-flash":{"mode":"video_generation","source":"https://replicate.com/wan-video/wan2.6-i2v-flash","provider":"replicate","base_model":"wan-video/wan2.6-i2v-flash"},"replicate/wavespeedai/qwen-image":{"mode":"image_generation","source":"https://replicate.com/wavespeedai/qwen-image","provider":"replicate","base_model":"wavespeedai/qwen-image"},"replicate/wavespeedai/wan-2.1-i2v-480p":{"mode":"video_generation","source":"https://replicate.com/wavespeedai/wan-2.1-i2v-480p","provider":"replicate","base_model":"wavespeedai/wan-2.1-i2v-480p"},"replicate/wavespeedai/wan-2.1-i2v-720p":{"mode":"video_generation","source":"https://replicate.com/wavespeedai/wan-2.1-i2v-720p","provider":"replicate","base_model":"wavespeedai/wan-2.1-i2v-720p"},"replicate/wavespeedai/wan-2.1-t2v-480p":{"mode":"video_generation","source":"https://replicate.com/wavespeedai/wan-2.1-t2v-480p","provider":"replicate","base_model":"wavespeedai/wan-2.1-t2v-480p"},"replicate/wavespeedai/wan-2.1-t2v-720p":{"mode":"video_generation","source":"https://replicate.com/wavespeedai/wan-2.1-t2v-720p","provider":"replicate","base_model":"wavespeedai/wan-2.1-t2v-720p"},"replicate/xai/grok-imagine-image":{"mode":"image_generation","provider":"replicate","base_model":"grok-imagine-image"},"replicate/xai/grok-imagine-image-quality":{"mode":"image_generation","source":"https://replicate.com/xai/grok-imagine-image-quality","provider":"replicate","base_model":"grok-imagine-image-quality"},"replicate/xai/grok-imagine-r2v":{"mode":"video_generation","source":"https://replicate.com/xai/grok-imagine-r2v","provider":"replicate","base_model":"grok-imagine-r2v"},"replicate/xai/grok-imagine-video":{"mode":"video_generation","source":"https://replicate.com/xai/grok-imagine-video","provider":"replicate","base_model":"grok-imagine-video"},"replicate/xai/grok-imagine-video-1.5":{"mode":"video_generation","source":"https://replicate.com/xai/grok-imagine-video-1.5","provider":"replicate","base_model":"grok-imagine-video-1.5"},"replicate/xai/grok-imagine-video-extension":{"mode":"video_generation","source":"https://replicate.com/xai/grok-imagine-video-extension","provider":"replicate","base_model":"grok-imagine-video-extension"},"replicate/xai/grok-text-to-speech":{"mode":"audio_speech","source":"https://replicate.com/xai/grok-text-to-speech","provider":"replicate","base_model":"grok-text-to-speech"},"fal_ai/blackforestlabs/flux-3/edit-image":{"metadata":{"notes":"the page states only the 1K (1 MP) price of $0.048 and the pricing API bills in megapixels at $0.048, so the per-pixel rate is $0.048 per megapixel and the per-image price is the default 1k output; 768sq, 1.5k, 2k and 4k outputs are priced by the per-pixel rate"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/edit-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"blackforestlabs/flux-3/edit-image"},"fal_ai/blackforestlabs/flux-3/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; the pricing API reports a $0.085 per second base price that the page does not state"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/image-to-video"},"fal_ai/blackforestlabs/flux-3/klein/edit-image":{"metadata":{"notes":"per output image by resolution; output_cost_per_image is the 1K rate (image_size defaults to the input size, 1K is the page example); 768 x 768 is $0.008 and 1.5K is $0.011 per image, no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/edit-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/edit-image"},"fal_ai/blackforestlabs/flux-3/klein/text-to-image":{"metadata":{"notes":"per output image by resolution; output_cost_per_image is the 1K rate (default image_size square_hd); 768 x 768 is $0.008 and 1.5K is $0.011 per image, no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/klein/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/klein/text-to-image"},"fal_ai/blackforestlabs/flux-3/text-to-image":{"metadata":{"notes":"the page states only the 1K (1 MP) price of $0.048 and the pricing API bills in megapixels at $0.048, so the per-pixel rate is $0.048 per megapixel and the per-image price is the default 1k output; 768sq, 1.5k, 2k and 4k outputs are priced by the per-pixel rate"},"mode":"image_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/text-to-image"},"fal_ai/blackforestlabs/flux-3/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; the pricing API reports a $0.085 per second base price that the page does not state"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-3/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-3/text-to-video"},"fal_ai/blackforestlabs/flux-video-upscale":{"metadata":{"comment":"per second of output video by resolution in the default creative mode (creativity=1); output_cost_per_second is the highest tier (4K) because the output resolution depends on the source and upscale_factor; precise mode (creativity=0) is $0.084 (1080p), $0.15 (2K) and $0.33 (4K) per second and has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/blackforestlabs/flux-video-upscale","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"blackforestlabs/flux-video-upscale"},"fal_ai/bria/fibo-2/generate":{"metadata":{"notes":"per generated image at the default 1MP output; the 4MP output is $0.06 per image and has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/bria/fibo-2/generate","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"bria/fibo-2/generate"},"fal_ai/bytedance/seedance-2.0/fast/reference-to-video":{"metadata":{"comment":"per second of generated video at the default resolution 720p, the only per-second rate the page states; fal bills tokens at $0.0112 per 1000 tokens, counts input video duration, and multiplies by 0.6 with video inputs"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/fast/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/fast/reference-to-video"},"fal_ai/bytedance/seedance-2.0/us/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; 1080p and 4k are not priced on the page; fal bills tokens at $0.0168 per 1000 tokens, counts input video duration, and multiplies by 0.6 with video inputs"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.0/us/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.0/us/reference-to-video"},"fal_ai/bytedance/seedance-2.5/us/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; with video references fal bills input and output duration and multiplies by 0.6"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/us/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/us/reference-to-video"},"fal_ai/fal-ai/bria/background/remove":{"metadata":{"notes":"per image without a prompt (prompt is optional); $0.03 per image when a prompt is given has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/bria/background/remove","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/background/remove"},"fal_ai/fal-ai/bytedance-upscaler/upscale/video":{"metadata":{"comment":"per second of output video by target_resolution at 30fps in normal mode; output_cost_per_second is the default target_resolution 1080p rate; 6k $0.0574 and 8k $0.1148 per second have no catalog field; 30-60fps doubles, over 60fps quadruples and pro mode multiplies by 10, none of which is modeled"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/bytedance-upscaler/upscale/video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"bytedance-upscaler/upscale/video"},"fal_ai/fal-ai/emu-3.5-image/edit-image":{"metadata":{"notes":"per image at the default 720p resolution; 480p is $0.15 per image and has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/emu-3.5-image/edit-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"emu-3.5-image/edit-image"},"fal_ai/fal-ai/emu-3.5-image/text-to-image":{"metadata":{"notes":"per image at the default 720p resolution; 480p is $0.15 per image and has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/emu-3.5-image/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"emu-3.5-image/text-to-image"},"fal_ai/fal-ai/kling-video/v2.5-turbo/pro/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.5-turbo/pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.5-turbo/pro/image-to-video"},"fal_ai/fal-ai/kling-video/v2.5-turbo/pro/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.5-turbo/pro/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.5-turbo/pro/text-to-video"},"fal_ai/fal-ai/kling-video/v2.5-turbo/standard/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.5-turbo/standard/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.5-turbo/standard/image-to-video"},"fal_ai/fal-ai/minimax/hailuo-02/standard/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 768P rate; 512P is roughly $0.017 per second and has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/minimax/hailuo-02/standard/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"hailuo-02/standard/image-to-video"},"fal_ai/fal-ai/minimax/image-01":{"mode":"image_generation","source":"https://fal.ai/models/fal-ai/minimax/image-01","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"image-01"},"fal_ai/fal-ai/object-removal":{"metadata":{"notes":"per image at the default model best_quality; low_quality $0.006, medium_quality $0.012 and high_quality $0.018 per image have no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/object-removal","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"object-removal"},"fal_ai/fal-ai/object-removal/bbox":{"metadata":{"notes":"per image at the default model best_quality; low_quality $0.006, medium_quality $0.012 and high_quality $0.018 per image have no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/object-removal/bbox","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"object-removal/bbox"},"fal_ai/fal-ai/object-removal/mask":{"metadata":{"notes":"per image at the default model best_quality; low_quality $0.006, medium_quality $0.012 and high_quality $0.018 per image have no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/object-removal/mask","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"object-removal/mask"},"fal_ai/fal-ai/recraft/v3/image-to-image":{"metadata":{"notes":"per image at the default realistic_image style; vector styles are $0.08 per image and have no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/recraft/v3/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"v3/image-to-image"},"fal_ai/fal-ai/scail-2":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/scail-2","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"scail-2"},"fal_ai/fal-ai/wan-vace-apps/long-reframe":{"metadata":{"comment":"per second of output video by resolution, converted from the page rate per 2 seconds ($0.08 per 2s at 720p, $0.04 per 2s at 480p); output_cost_per_second is the 720p rate because the default resolution is auto; 580p ($0.06 per 2s) has no catalog field, 240p and 360p are not priced on the page; seconds are rounded up to a multiple of 2"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan-vace-apps/long-reframe","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan-vace-apps/long-reframe"},"fal_ai/fal-ai/wan/v2.2-14b/animate/move":{"metadata":{"comment":"per video second of output by resolution, where a video second is 16 frames; output_cost_per_second is the default resolution 480p rate; 580p is $0.06 per video second and has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-14b/animate/move","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-14b/animate/move"},"fal_ai/fal-ai/wan/v2.2-14b/animate/replace":{"metadata":{"comment":"per video second of output by resolution, where a video second is 16 frames; output_cost_per_second is the default resolution 480p rate; 580p is $0.06 per video second and has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/wan/v2.2-14b/animate/replace","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"wan/v2.2-14b/animate/replace"},"fal_ai/google/gemini-omni-flash/v1.2/edit":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; 360p is $0.0273 per second and has no catalog field; input tokens (text, image, video) are billed at $1.50 per million"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/edit","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/edit"},"fal_ai/google/gemini-omni-flash/v1.2/extend":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; 360p is $0.0273 per second and has no catalog field; input tokens (text, image, video) are billed at $1.50 per million"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/extend","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/extend"},"fal_ai/google/gemini-omni-flash/v1.2/image-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; 360p is $0.0273 per second and has no catalog field; input tokens (text, image, video) are billed at $1.50 per million"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/image-to-video"},"fal_ai/google/gemini-omni-flash/v1.2/reference-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; 360p is $0.0273 per second and has no catalog field; input tokens (text, image, video) are billed at $1.50 per million"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/reference-to-video"},"fal_ai/google/gemini-omni-flash/v1.2/text-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; 360p is $0.0273 per second and has no catalog field; input tokens (text, image, video) are billed at $1.50 per million"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/text-to-video"},"fal_ai/google/virtual-try-on":{"mode":"image_generation","source":"https://fal.ai/models/google/virtual-try-on","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"virtual-try-on"},"fal_ai/ideogram/v4/tiling":{"metadata":{"notes":"per megapixel at the default rendering_speed BALANCED; TURBO is $0.03 and QUALITY is $0.10 per megapixel, no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/ideogram/v4/tiling","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v4/tiling"},"fal_ai/ideogram/v4/tiling/lora":{"metadata":{"notes":"per megapixel at the default rendering_speed BALANCED; TURBO is $0.03375 and QUALITY is $0.1125 per megapixel, no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/ideogram/v4/tiling/lora","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"ideogram/v4/tiling/lora"},"fal_ai/kling/v4/flash/reference-to-video":{"metadata":{"comment":"per second of output video at 720p (the only resolution) with image references; with video references fal bills $0.063 per second on the output duration plus the greater of the input and output durations, which has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/flash/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/flash/reference-to-video"},"fal_ai/kling/v4/reference-to-video":{"metadata":{"comment":"per second of output video by resolution with image references; output_cost_per_second is the default resolution 1080p rate; with video references or video elements fal bills $0.094 (720p) or $0.141 (1080p) per second on the input plus output length, which has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/kling/v4/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling/v4/reference-to-video"},"fal_ai/luma/agent/ray/v3.2/reframe":{"metadata":{"comment":"per started source second by resolution; output_cost_per_second is the default resolution 540p rate ($0.06), which has no tier field"},"mode":"video_generation","source":"https://fal.ai/models/luma/agent/ray/v3.2/reframe","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"luma/agent/ray/v3.2/reframe"},"fal_ai/minimax/h3-max/reference-to-video":{"metadata":{"comment":"per second of requested output duration by resolution; output_cost_per_second is the default resolution 768P rate; the page states no promotion for this endpoint; reference tokens above the 4,096 included per request are $0.02 per 1,000 tokens and are not modeled"},"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/reference-to-video"},"fal_ai/minimax/h3-max/styles/16bit-pixel":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/styles/16bit-pixel","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/styles/16bit-pixel"},"fal_ai/minimax/h3-max/styles/hand-drawn":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/styles/hand-drawn","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/styles/hand-drawn"},"fal_ai/minimax/h3-max/styles/low-poly":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/styles/low-poly","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/styles/low-poly"},"fal_ai/minimax/h3-max/styles/retro-toon-70s":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/styles/retro-toon-70s","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/styles/retro-toon-70s"},"fal_ai/minimax/h3-max/styles/vhs":{"mode":"video_generation","source":"https://fal.ai/models/minimax/h3-max/styles/vhs","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"h3-max/styles/vhs"},"fal_ai/mirelo-ai/sfx2.0/video-to-video":{"metadata":{"comment":"per second of generated audio per sample; promotional 33% off rate through 2026-12-13 per the page; the standard rate of $0.015 per second applies from 2026-12-14 (ruling fal-ai-promotional-rates-apply-until-promotion-ends, 2026-11-10); mode and endpoint follow fal_ai/sonilo/v1.1/video-to-video-sound-effects"},"mode":"video_generation","source":"https://fal.ai/models/mirelo-ai/sfx2.0/video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"mirelo-ai/sfx2.0/video-to-video"},"fal_ai/nvidia/cosmos-3-super/text-to-image":{"metadata":{"notes":"per generated image; the optional $0.02 per request prompt expansion surcharge is not modeled, and agentic generation bills every candidate image"},"mode":"image_generation","source":"https://fal.ai/models/nvidia/cosmos-3-super/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"cosmos-3-super/text-to-image"},"fal_ai/recraft/v4/style/pro/text-to-image":{"metadata":{"notes":"per generated image when a style_id is provided; the one-time $0.005 style creation surcharge when no style_id is given is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/recraft/v4/style/pro/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/style/pro/text-to-image"},"fal_ai/recraft/v4/style/pro/text-to-vector":{"metadata":{"notes":"per generated image when a style_id is provided; the one-time $0.005 style creation surcharge when no style_id is given is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/recraft/v4/style/pro/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/style/pro/text-to-vector"},"fal_ai/recraft/v4/style/text-to-image":{"metadata":{"notes":"per generated image when a style_id is provided; the one-time $0.005 style creation surcharge when no style_id is given is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/recraft/v4/style/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/style/text-to-image"},"fal_ai/recraft/v4/style/text-to-vector":{"metadata":{"notes":"per generated image when a style_id is provided; the one-time $0.005 style creation surcharge when no style_id is given is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/recraft/v4/style/text-to-vector","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"v4/style/text-to-vector"},"fal_ai/vast/eden-live-14b-video":{"metadata":{"comment":"per second of video generated in a live session; the 10 billable second ($0.10) minimum per session is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/vast/eden-live-14b-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vast/eden-live-14b-video"},"fal_ai/vast/eden-live-5b-video":{"metadata":{"comment":"per second of video generated in a live session; the 10 billable second ($0.035) minimum per session is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/vast/eden-live-5b-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vast/eden-live-5b-video"},"fal_ai/xai/grok-imagine-image/edit":{"mode":"image_generation","source":"https://fal.ai/models/xai/grok-imagine-image/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"grok-imagine-image/edit"},"fal_ai/xai/grok-imagine-video/edit-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the 720p rate because the default resolution is auto; the $0.01 per second of input video charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/edit-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/edit-video"},"fal_ai/xai/grok-imagine-video/extend-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the 720p rate because the endpoint has no resolution field; the $0.01 per second of input video charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/extend-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/extend-video"},"fal_ai/xai/grok-imagine-video/image-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; the $0.002 per request image input charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/image-to-video"},"fal_ai/xai/grok-imagine-video/reference-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 480p rate; the $0.002 per request image input charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/reference-to-video"},"fal_ai/xai/grok-imagine-video/text-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/text-to-video"},"fal_ai/xai/grok-imagine-video/v1.5/image-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 720p rate; the $0.01 per reference image charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/v1.5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/v1.5/image-to-video"},"fal_ai/xai/grok-imagine-video/v1.5/reference-to-video":{"metadata":{"comment":"per second of output video by resolution; output_cost_per_second is the default resolution 480p rate; the $0.01 per reference image charge is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/xai/grok-imagine-video/v1.5/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"grok-imagine-video/v1.5/reference-to-video"},"databricks/databricks-claude-fable-5-2":{"max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_anthropic_thinking_payload":true,"supports_assistant_prefill":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"databricks","base_model":"claude-fable-5-2"},"databricks/system.ai.claude-fable-5-2":{"max_input_tokens":1000000,"max_output_tokens":128000,"metadata":{"notes":"Claude Fable 5.2 on Databricks is served through Unity Gateway with the model service name system.ai.claude-fable-5-2, hosted through Amazon Bedrock. Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},"mode":"chat","prompt_cache_min_tokens":512,"source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_adaptive_thinking":true,"supports_anthropic_thinking_payload":true,"supports_assistant_prefill":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_mid_conversation_system":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"thinking_always_on":true,"provider":"databricks","base_model":"system.ai.claude-fable-5-2"},"replicate/google/gemini-4-carbon":{"max_input_tokens":1048576,"max_output_tokens":65535,"mode":"chat","source":"https://replicate.com/google/gemini-4-carbon","supports_audio_input":true,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_reasoning":true,"supports_response_schema":false,"supports_system_messages":true,"supports_tool_choice":false,"supports_video_input":true,"supports_vision":true,"provider":"replicate","base_model":"gemini-4-carbon"},"azure/mistral-large-4-latest":{"deprecation_date":"2027-10-27","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://docs.mistral.ai/models/mistral-large-4-0","supports_assistant_prefill":true,"supports_function_calling":true,"supports_pdf_input":false,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"azure","base_model":"mistral-large-4"},"fal_ai/fal-ai/recraft-20b":{"metadata":{"notes":"per image at the default realistic_image style, from the fal pricing API; vector styles cost 2x and have no catalog field; the fal.ai page for this id redirects to fal-ai/recraft/v2/text-to-image, which rounds the rate to $0.022"},"mode":"image_generation","source":"https://api.fal.ai/v1/models/pricing?endpoint_id=fal-ai/recraft-20b","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"recraft-20b"},"fal_ai/fal-ai/stable-diffusion-v3-medium/image-to-image":{"mode":"image_generation","source":"https://api.fal.ai/v1/models/pricing?endpoint_id=fal-ai/stable-diffusion-v3-medium/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"stable-diffusion-v3-medium/image-to-image"},"fal_ai/fal-ai/pixverse/lipsync":{"metadata":{"comment":"per second of output video; when audio_url is not provided the text is spoken with TTS at $0.24 per 100 characters, which is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/lipsync","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/lipsync"},"bedrock_mantle/us-gov-east-1/openai.gpt-5.6-cyber":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-cyber.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"use_openai_responses_path":true,"provider":"bedrock_mantle","base_model":"gpt-5.6-cyber"},"fal_ai/fal-ai/kling-video/v2.6/pro/image-to-video":{"metadata":{"comment":"per second of generated video with the default generate_audio=true ($0.14); videos without audio are $0.07 per second and videos with audio and voice control are $0.168 per second, which have no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/kling-video/v2.6/pro/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"kling-video/v2.6/pro/image-to-video"},"fal_ai/fal-ai/vidu/image-to-video":{"metadata":{"comment":"per second derived from the fixed 4s clip price of $0.20 per video"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/image-to-video"},"fal_ai/fal-ai/vidu/start-end-to-video":{"metadata":{"comment":"per second derived from the fixed 4s clip price of $0.20 per video"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/start-end-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/start-end-to-video"},"fal_ai/fal-ai/vidu/reference-to-video":{"metadata":{"comment":"per second derived from the fixed 4s clip price of $0.40 per video"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/reference-to-video"},"fal_ai/luma/agent/ray/v3.2/image-to-video":{"metadata":{"comment":"per second derived from the 5s clip prices by resolution ($0.15 at 540p, $0.30 at 720p, $1.20 at 1080p); output_cost_per_second is the default resolution 540p rate, which has no tier field; HDR ($0.60 at 720p, $2.40 at 1080p per 5s) and HDR+EXR ($0.90, $3.60) are not modeled"},"mode":"video_generation","source":"https://fal.ai/models/luma/agent/ray/v3.2/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"luma/agent/ray/v3.2/image-to-video"},"fal_ai/luma/agent/ray/v3.2/text-to-video":{"metadata":{"comment":"per second derived from the 5s and 10s clip prices by resolution ($0.50/$1.00 at 540p, $1.00/$2.00 at 720p, $2.00/$4.00 at 1080p); output_cost_per_second is the default resolution 540p rate, which has no tier field"},"mode":"video_generation","source":"https://fal.ai/models/luma/agent/ray/v3.2/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"luma/agent/ray/v3.2/text-to-video"},"fal_ai/luma/agent/ray/v3.2/video-to-video":{"metadata":{"comment":"per second derived from the 5s and 10s clip prices by resolution ($0.72/$1.44 at 540p, $1.08/$2.16 at 720p, $2.16/$4.32 at 1080p); output_cost_per_second is the default resolution 540p rate, which has no tier field; HDR and HDR+EXR prices are not modeled"},"mode":"video_generation","source":"https://fal.ai/models/luma/agent/ray/v3.2/video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"luma/agent/ray/v3.2/video-to-video"},"fal_ai/pixverse/v6/max/text-to-video":{"metadata":{"comment":"per second of generated video by resolution without audio (the default generate_audio_switch=false); output_cost_per_second is the default resolution 720p rate; 360p $0.054 and 540p $0.081 have no catalog field; with audio the rates are $0.081/$0.099/$0.135/$0.252 per second, which is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/pixverse/v6/max/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v6/max/text-to-video"},"fal_ai/pixverse/v6/max/image-to-video":{"metadata":{"comment":"per second of generated video by resolution without audio (the default generate_audio_switch=false); output_cost_per_second is the default resolution 720p rate; 360p $0.054 and 540p $0.081 have no catalog field; with audio the rates are $0.081/$0.099/$0.135/$0.252 per second, which is not modeled"},"mode":"video_generation","source":"https://fal.ai/models/pixverse/v6/max/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v6/max/image-to-video"},"fal_ai/fal-ai/flux-2-pro":{"metadata":{"comment":"per image for the default 1 MP output; fal bills $0.03 for the first output megapixel plus $0.015 per additional input or output megapixel, rounded up, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-pro","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-pro"},"fal_ai/fal-ai/flux-2-pro/edit":{"metadata":{"comment":"per image for the default 1 MP output; fal bills $0.03 for the first output megapixel plus $0.015 per additional input or output megapixel, rounded up, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-pro/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-pro/edit"},"fal_ai/fal-ai/flux-2-pro/outpaint":{"metadata":{"comment":"per image for the default 1 MP output; fal bills $0.03 for the first output megapixel plus $0.015 per additional input or output megapixel, rounded up, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-pro/outpaint","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-pro/outpaint"},"fal_ai/fal-ai/flux-2-max":{"metadata":{"comment":"per image for the default 1 MP output; fal bills $0.07 for the first processed megapixel plus $0.03 per additional megapixel, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-max","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"flux-2-max"},"fal_ai/fal-ai/flux-2-max/edit":{"metadata":{"comment":"per image for the default 1 MP output; fal bills $0.07 for the first processed megapixel plus $0.03 per additional megapixel, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-2-max/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-2-max/edit"},"fal_ai/fal-ai/flux-pro/v1/erase":{"metadata":{"comment":"per image for the first generated megapixel; fal bills $0.004 for each subsequent generated megapixel and $0.004 per reference megapixel with a 3 MP minimum, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1/erase","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1/erase"},"fal_ai/fal-ai/smart-resize":{"metadata":{"comment":"per single-image request at standard resolution: $0.15 per output image plus the $0.05 minimum vision analysis fee per request; additional output images in one request ($0.15 each) and 4K outputs (charged double) are not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/smart-resize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"smart-resize"},"fal_ai/bytedance/seedream/v5/pro/layerize":{"metadata":{"comment":"per generated layer for total pixel area under 1536x1536; layers for a base image over 1536x1536 are $0.0675 each, which has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/pro/layerize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5/pro/layerize"},"fal_ai/luma/agent/uni-1/v1/edit":{"metadata":{"comment":"per edit with one source image ($0.042 per edit plus $0.003 per reference or source image); each additional reference image adds $0.003, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/luma/agent/uni-1/v1/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"luma/agent/uni-1/v1/edit"},"fal_ai/luma/agent/uni-1/v1/max/edit":{"metadata":{"comment":"per edit with one source image ($0.102 per edit plus $0.003 per reference or source image); each additional reference image adds $0.003, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/luma/agent/uni-1/v1/max/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"luma/agent/uni-1/v1/max/edit"},"databricks/databricks-gemini-4-carbon":{"max_input_tokens":1048576,"max_output_tokens":1048576,"metadata":{"notes":"Gemini 4 Carbon on Databricks is available only through Unity Gateway with the model service name system.ai.gemini-4-carbon on a global endpoint. The Databricks pricing pages list no row for it yet, so input and output are the Global rates on the Vertex AI pricing page for Gemini 4 Carbon (listed there through November 30, 2026), no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},"mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"databricks","base_model":"gemini-4-carbon"},"databricks/system.ai.gemini-4-carbon":{"max_input_tokens":1048576,"max_output_tokens":1048576,"metadata":{"notes":"Gemini 4 Carbon on Databricks is available only through Unity Gateway with the model service name system.ai.gemini-4-carbon on a global endpoint. The Databricks pricing pages list no row for it yet, so input and output are the Global rates on the Vertex AI pricing page for Gemini 4 Carbon (listed there through November 30, 2026), no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},"mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"provider":"databricks","base_model":"system.ai.gemini-4-carbon"},"runwayml/kling2.5_turbo_pro":{"metadata":{"comment":"12 credits per second @ $0.01 per credit = $0.12 per second"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling2.5-turbo-pro"},"runwayml/kling3.0_standard":{"metadata":{"comment":"13 credits per second with audio, 9 without @ $0.01 per credit; priced at the with-audio rate (audio defaults to true)"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling3.0-standard"},"runwayml/kling3.0_pro":{"metadata":{"comment":"17 credits per second with audio, 12 without @ $0.01 per credit; priced at the with-audio rate (audio defaults to true)"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling3.0-pro"},"runwayml/kling3.0_4k":{"metadata":{"comment":"43 credits per second @ $0.01 per credit = $0.43 per second, with or without audio"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling3.0-4k"},"runwayml/klingO3_standard":{"metadata":{"comment":"13 credits per second @ $0.01 per credit = $0.13 per second; video-to-video is billed per second of the input video, not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"klingo3-standard"},"runwayml/klingO3_pro":{"metadata":{"comment":"17 credits per second @ $0.01 per credit = $0.17 per second; video-to-video is billed per second of the input video, not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"klingo3-pro"},"runwayml/klingO3_4k":{"metadata":{"comment":"43 credits per second @ $0.01 per credit = $0.43 per second; video-to-video is billed per second of the input video, not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"klingo3-4k"},"runwayml/kling_4_flash":{"metadata":{"comment":"9.4 credits per second (720p) without input video @ $0.01 per credit = $0.094 per second; the with-input-video rate (6.3 credits per output second plus 6.3 per second of the greater of input and output duration), rounding up to the next credit and the 29 credit minimum are not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling-4-flash"},"runwayml/wan3":{"metadata":{"comment":"5/10/20 credits per second at 480p/720p/1080p @ $0.01 per credit"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"wan3"},"runwayml/wan3_prime":{"metadata":{"comment":"6.8/14/28 credits per second at 480p/720p/1080p @ $0.01 per credit"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"wan3-prime"},"runwayml/grok_imagine_1_5":{"metadata":{"comment":"10/16/29 credits per second at 480p/720p/1080p @ $0.01 per credit; 1 credit per image or audio reference not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"grok-imagine-1-5"},"runwayml/grok_imagine_1_5_lite":{"metadata":{"comment":"2/3/14 credits per second at 480p/720p/1080p @ $0.01 per credit; 1 credit for an image-to-video start frame not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"grok-imagine-1-5-lite"},"runwayml/h3_max":{"metadata":{"comment":"8 credits per second at 768p, 5 at 480p @ $0.01 per credit"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"h3-max"},"runwayml/happyhorse_1_0":{"metadata":{"comment":"15 credits per second at 720p, 30 at 1080p @ $0.01 per credit"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"happyhorse-1-0"},"runwayml/gemini_omni_flash_1.2":{"metadata":{"comment":"7/10/21 credits per second at 720p/1080p/4K @ $0.01 per credit; the 360p rate (2 credits per second), 1 credit per reference image or keyframe, 1 credit per second of input video and the 6 credit minimum per request are not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"gemini-omni-flash-1.2"},"runwayml/gemini_omni_flash_1.1":{"metadata":{"comment":"10/15/30 credits per second at 720p/1080p/4K @ $0.01 per credit; the 360p rate (3.4 credits per second), 1 credit per reference image and 1 credit per second of input video are not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"gemini-omni-flash-1.1"},"runwayml/muse_image":{"metadata":{"comment":"1 credit per image (any resolution) @ $0.01 per credit = $0.01 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"muse-image"},"runwayml/seedream5_lite":{"metadata":{"comment":"4 credits per image (any resolution) @ $0.01 per credit = $0.04 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"seedream5-lite"},"runwayml/seedream5_flash":{"metadata":{"comment":"2 credits per image (any resolution) @ $0.01 per credit = $0.02 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"seedream5-flash"},"runwayml/gemini_image3.1_flash_lite":{"metadata":{"comment":"4 credits per 1K image @ $0.01 per credit = $0.04 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"gemini-image3.1-flash-lite"},"runwayml/bria_fibo_edit_1_5":{"metadata":{"comment":"5 credits per image @ $0.01 per credit = $0.05 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"bria-fibo-edit-1-5"},"runwayml/gemini_2.5_flash":{"metadata":{"comment":"5 credits per image (any resolution) @ $0.01 per credit = $0.05 per image"},"mode":"image_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image"],"supported_output_modalities":["image"],"provider":"runwayml","base_model":"gemini-2.5-flash"},"runwayml/eleven_v3":{"metadata":{"comment":"1 credit per 50 characters @ $0.01 per credit = $0.0002 per character; 1 credit minimum not modeled"},"mode":"audio_speech","source":"https://docs.dev.runwayml.com/guides/pricing/","provider":"runwayml","base_model":"eleven-v3"},"runwayml/eleven_v4":{"metadata":{"comment":"5 credits per 1,000 characters @ $0.01 per credit = $0.00005 per character; 1 credit minimum not modeled"},"mode":"audio_speech","source":"https://docs.dev.runwayml.com/guides/pricing/","provider":"runwayml","base_model":"eleven-v4"},"gemini/gemini-decision-1":{"max_input_tokens":262144,"mode":"decisions","rpm":1500,"source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/decisions"],"supported_modalities":["text","image","audio","video"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":false,"supports_function_calling":false,"supports_parallel_function_calling":false,"supports_pdf_input":true,"supports_prompt_caching":false,"supports_reasoning":false,"supports_response_schema":false,"supports_tool_choice":false,"supports_url_context":false,"supports_video_input":true,"supports_vision":true,"supports_web_search":false,"tpm":1000000,"provider":"gemini","base_model":"gemini-decision-1"},"runwayml/kling_4":{"metadata":{"comment":"Promo rate through 2026-11-30 UTC: 12.56 credits per second (720p) and 18.8 (1080p) without input video @ $0.01 per credit = $0.1256 and $0.188 per second; list rate 15.7/23.5 credits per second applies from 2026-12-01 UTC (ruling runwayml-kling-4-promo-rate). The with-input-video rate (7.52/11.28 credits per output second plus the same per second of the greater of input and output duration) and rounding up to the next credit are not modeled"},"mode":"video_generation","source":"https://docs.dev.runwayml.com/guides/pricing/","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"provider":"runwayml","base_model":"kling-4"},"fal_ai/bytedance/seedream/v5.5/text-to-image":{"metadata":{"comment":"per image at the default image_size auto_2K; 1K and 1.5K outputs are $0.03 and custom dimensions are $0.03 up to and including 2,610,000 pixels and $0.0405 above, so 1.5K has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5.5/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v5.5/text-to-image"},"fal_ai/bytedance/seedream/v5.5/edit":{"metadata":{"comment":"per output image with one input image at the default image_size auto_2K ($0.0405 per output plus $0.012 per input image per output); 1K and 1.5K outputs are $0.03 plus the input fee and each additional input image adds $0.012 per output, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5.5/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5.5/edit"},"fal_ai/bytedance/seedream/v5.5/layerize":{"metadata":{"comment":"per returned output image (the base image and every separated layer) at the default image_size auto, where a base image above 2,610,000 pixels bills $0.0405 per output and one at or under that area bills $0.03; fal also bills $0.012 for the input image once per request and returns up to max_images (default 30) outputs, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5.5/layerize","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5.5/layerize"},"fal_ai/fal-ai/flux-pro/v1/vto":{"metadata":{"comment":"per try-on with one 1 MP input image and one 1 MP output ($0.0375 for the first input megapixel plus $0.005 per output megapixel); fal bills $0.005 per additional input megapixel and rounds each input image and the output up to whole megapixels separately, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/flux-pro/v1/vto","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"flux-pro/v1/vto"},"fal_ai/fal-ai/patina/material":{"metadata":{"comment":"per request at the schema defaults (1024x1024 square_hd, all 5 maps, no upscaling), the page's own example; fal bills $0.01 per request plus $0.02 per megapixel plus $0.01 per megapixel per map type, and 2x or 4x upscaling adds $0.004 or $0.016 per pre-upscaling megapixel per map, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/patina/material","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"patina/material"},"fal_ai/fal-ai/patina/material/extract":{"metadata":{"comment":"per request at the schema defaults (1024x1024 input, all 5 maps, no upscaling), the page's own example; fal bills $0.10 per request plus $0.02 per megapixel plus $0.01 per megapixel per map type, and 2x or 4x upscaling adds $0.004 or $0.016 per pre-upscaling megapixel per map, which is not modeled"},"mode":"image_generation","source":"https://fal.ai/models/fal-ai/patina/material/extract","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"patina/material/extract"},"fal_ai/fal-ai/vidu/template-to-video":{"metadata":{"comment":"per second derived from the $0.20 Standard template price over the fixed 4s clip the sibling vidu endpoints state; Premium templates are $0.30 and Advanced templates $0.50 per video, which has no catalog field"},"mode":"video_generation","source":"https://fal.ai/models/fal-ai/vidu/template-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"vidu/template-to-video"},"gemini/gemini-4-hydrogen":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","prompt_cache_min_tokens":4096,"rpm":2000,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://aistudio.google.com/docs/pricing#gemini-4-hydrogen","supported_endpoints":["/v1/chat/completions","/v1/completions"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"tpm":800000,"web_search_billing_unit":"per_query","provider":"gemini","base_model":"gemini-4-hydrogen"},"vertex_ai/gemini-4-hydrogen":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","prompt_cache_min_tokens":4096,"regional_endpoint_uplift_multiplier":1.1,"search_free_requests":{"count":5000,"period":"month","shared_group":"gemini-3"},"source":"https://cloud.google.com/gemini-enterprise-agent-platform/generative-ai/pricing","supported_endpoints":["/v1/chat/completions","/v1/completions","/v1/batch"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_audio_output":false,"supports_computer_use":true,"supports_continuation":true,"supports_function_calling":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_url_context":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"web_search_billing_unit":"per_query","provider":"vertex_ai","base_model":"gemini-4-hydrogen"},"together_ai/thinkingmachines/Inkling-2-Small":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","source":"https://api.together.ai/v1/models","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"provider":"together_ai","base_model":"thinkingmachines/inkling-2-small"},"fal_ai/google/gemini-omni-flash/v1.2/upscale":{"metadata":{"comment":"per second of upscaled output video by target resolution; the 1080p tier and output_cost_per_second take the higher source-dependent rate ($0.0819 from a 360p source) because the catalog selects by target resolution alone and the source resolution has no field, so a 720p source upscaled to 1080p ($0.041 per second) is overcounted, the same convention as fal_ai/blackforestlabs/flux-video-upscale"},"mode":"video_generation","source":"https://fal.ai/models/google/gemini-omni-flash/v1.2/upscale","supported_endpoints":["/v1/videos"],"supported_modalities":["video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"gemini-omni-flash/v1.2/upscale"},"fal_ai/decart/lucy-5b/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/decart/lucy-5b/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"decart/lucy-5b/image-to-video"},"fal_ai/fal-ai/decart/lucy-5b/image-to-video":{"alias_of":"fal_ai/decart/lucy-5b/image-to-video","base_model":"decart/lucy-5b/image-to-video"},"github_copilot/gemini-4-hydrogen":{"max_input_tokens":200000,"max_output_tokens":65536,"mode":"chat","source":"https://github.blog/changelog/2026-11-13-gemini-4-hydrogen-in-github-copilot","supported_endpoints":["/v1/chat/completions"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"github_copilot","base_model":"gemini-4-hydrogen"},"fal_ai/bytedance/seedream/v5/pro/us/edit":{"metadata":{"comment":"per output image by total pixel area at the default image_size auto_2K with one input image; outputs up to and including 1536x1536 pixels are $0.0675 and outputs above that up to 2048x2048 are $0.135; the first input image is free and each additional input image adds $0.0045 per output, which is not modeled; fal calls this pricing tentative"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/pro/us/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"seedream/v5/pro/us/edit"},"fal_ai/bytedance/seedream/v5/pro/us/text-to-image":{"metadata":{"comment":"per output image by total pixel area at the default image_size auto_2K; outputs up to and including 1536x1536 pixels are $0.0675 and outputs above that up to 2048x2048 are $0.135, so 1.5K has no catalog field"},"mode":"image_generation","source":"https://fal.ai/models/bytedance/seedream/v5/pro/us/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"seedream/v5/pro/us/text-to-image"},"fal_ai/bytedance/seedance-2.5/flash/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/flash/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/flash/image-to-video"},"fal_ai/bytedance/seedance-2.5/flash/reference-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate; with video references fal bills input and output duration and multiplies by 0.614, about 0.1901 per second at 720p, not modeled"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/flash/reference-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/flash/reference-to-video"},"fal_ai/bytedance/seedance-2.5/flash/text-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 720p rate"},"mode":"video_generation","source":"https://fal.ai/models/bytedance/seedance-2.5/flash/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"seedance-2.5/flash/text-to-video"},"cloudflare/@cf/thinkingmachines/inkling-2-small":{"max_input_tokens":1048576,"mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/inkling-2-small/","supports_function_calling":true,"supports_reasoning":true,"supports_vision":true,"provider":"cloudflare","base_model":""},"nebius/thinkingmachines/Inkling-2-Small":{"max_input_tokens":1048576,"mode":"chat","supports_vision":true,"supports_reasoning":true,"supports_function_calling":true,"source":"https://tokenfactory.nebius.com/api/public/models_info","provider":"nebius","base_model":"thinkingmachines/inkling-2-small"},"fal_ai/fal-ai/pixverse/extend":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/extend","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/extend"},"fal_ai/fal-ai/pixverse/extend/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/extend/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/extend/fast"},"fal_ai/fal-ai/pixverse/v3.5/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/effects"},"fal_ai/fal-ai/pixverse/v3.5/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/image-to-video"},"fal_ai/fal-ai/pixverse/v3.5/image-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/image-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/image-to-video/fast"},"fal_ai/fal-ai/pixverse/v3.5/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/text-to-video"},"fal_ai/fal-ai/pixverse/v3.5/text-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/text-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/text-to-video/fast"},"fal_ai/fal-ai/pixverse/v3.5/transition":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v3.5/transition","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v3.5/transition"},"fal_ai/fal-ai/pixverse/v4.5/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/effects"},"fal_ai/fal-ai/pixverse/v4.5/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/image-to-video"},"fal_ai/fal-ai/pixverse/v4.5/image-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/image-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/image-to-video/fast"},"fal_ai/fal-ai/pixverse/v4.5/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/text-to-video"},"fal_ai/fal-ai/pixverse/v4.5/text-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/text-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/text-to-video/fast"},"fal_ai/fal-ai/pixverse/v4.5/transition":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4.5/transition","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4.5/transition"},"fal_ai/fal-ai/pixverse/v4/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4/effects"},"fal_ai/fal-ai/pixverse/v4/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4/image-to-video"},"fal_ai/fal-ai/pixverse/v4/image-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4/image-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4/image-to-video/fast"},"fal_ai/fal-ai/pixverse/v4/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4/text-to-video"},"fal_ai/fal-ai/pixverse/v4/text-to-video/fast":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v4/text-to-video/fast","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v4/text-to-video/fast"},"fal_ai/fal-ai/pixverse/v5/effects":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v5/effects","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v5/effects"},"fal_ai/fal-ai/pixverse/v5/image-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v5/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v5/image-to-video"},"fal_ai/fal-ai/pixverse/v5/text-to-video":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v5/text-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v5/text-to-video"},"fal_ai/fal-ai/pixverse/v5/transition":{"mode":"video_generation","source":"https://fal.ai/models/fal-ai/pixverse/v5/transition","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"pixverse/v5/transition"},"fal_ai/bria/rmbg-3":{"mode":"image_generation","source":"https://fal.ai/models/bria/rmbg-3","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"bria/rmbg-3"},"fal_ai/speridlabs/iris-z/depth-to-image":{"metadata":{"notes":"fal bills speridlabs/iris-z/depth-to-image at $0.14 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/depth-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"speridlabs/iris-z/depth-to-image"},"fal_ai/speridlabs/iris-z/small/depth-to-image":{"metadata":{"notes":"fal bills speridlabs/iris-z/small/depth-to-image at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/small/depth-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"speridlabs/iris-z/small/depth-to-image"},"fal_ai/speridlabs/iris-z/small/text-to-image":{"metadata":{"notes":"fal bills speridlabs/iris-z/small/text-to-image at $0.03 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/small/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"speridlabs/iris-z/small/text-to-image"},"fal_ai/speridlabs/iris-z/small/text-to-rgbd":{"metadata":{"notes":"fal bills speridlabs/iris-z/small/text-to-rgbd at $0.05 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/small/text-to-rgbd","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"speridlabs/iris-z/small/text-to-rgbd"},"fal_ai/speridlabs/iris-z/text-to-image":{"metadata":{"notes":"fal bills speridlabs/iris-z/text-to-image at $0.1 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"speridlabs/iris-z/text-to-image"},"fal_ai/speridlabs/iris-z/text-to-rgbd":{"metadata":{"notes":"fal bills speridlabs/iris-z/text-to-rgbd at $0.14 per megapixel, rounding each image up to the nearest megapixel. The per-pixel rate is used when Fal reports the output size, and the flat per-image price prices the default 1 MP output like the sibling flux entries"},"mode":"image_generation","source":"https://fal.ai/models/speridlabs/iris-z/text-to-rgbd","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"speridlabs/iris-z/text-to-rgbd"},"omni-moderation-csam-latest":{"max_input_tokens":32768,"max_output_tokens":0,"mode":"moderation","source":"https://developers.openai.com/api/docs/guides/csam-classifier","provider":"openai","base_model":"omni-moderation-csam","web_search_billing_unit":"per_request"},"bedrock_mantle/anthropic.claude-fable-5":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"token_count_model":{"model":"anthropic.claude-fable-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-fable-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock_mantle/anthropic.claude-mythos-5":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"token_count_model":{"model":"anthropic.claude-mythos-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-mythos-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1"],"max_tokens":128000},"bedrock_mantle/anthropic.claude-opus-4-7":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-opus-4-7","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock_mantle/anthropic.claude-opus-4-8":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-opus-4-8","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock_mantle/anthropic.claude-opus-5":{"bedrock_converse_supports_strict_tools":false,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_on_by_default":true,"token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-opus-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock_mantle/anthropic.claude-sonnet-5":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_on_by_default":true,"token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock_mantle","base_model":"claude-sonnet-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-fable-5":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"token_count_model":{"model":"anthropic.claude-fable-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-fable-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-mythos-5":{"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"token_count_model":{"model":"anthropic.claude-mythos-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-mythos-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-opus-4-7":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":2048,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-7","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-opus-4-7","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-opus-4-8":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"token_count_model":{"model":"anthropic.claude-opus-4-8","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-opus-4-8","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-opus-5":{"bedrock_converse_supports_strict_tools":false,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_on_by_default":true,"token_count_model":{"model":"anthropic.claude-opus-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-opus-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-sonnet-5":{"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":1024,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_on_by_default":true,"token_count_model":{"model":"anthropic.claude-sonnet-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-sonnet-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-haiku-4-5":{"max_input_tokens":200000,"max_output_tokens":64000,"mode":"chat","prompt_cache_min_tokens":4096,"source":"https://aws.amazon.com/bedrock/pricing/","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"token_count_model":{"model":"anthropic.claude-haiku-4-5-20251001-v1:0","litellm_provider":"bedrock"},"provider":"bedrock","base_model":"claude-haiku-4-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":64000},"bedrock/mantle/anthropic.claude-opus-5-5":{"bedrock_converse_supports_strict_tools":false,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","supports_adaptive_thinking":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_mid_conversation_system":true,"supports_native_structured_output":false,"supports_output_config":true,"supports_parallel_tool_use_config":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"token_count_model":{"model":"anthropic.claude-opus-5-5","litellm_provider":"bedrock_mantle","region":"us-east-1"},"provider":"bedrock","base_model":"claude-opus-5-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-sonnet-5-5":{"bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":false,"thinking_always_on":true,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"claude-sonnet-5-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"bedrock/mantle/anthropic.claude-haiku-5-5":{"bedrock_converse_supports_strict_tools":false,"supports_tool_search":true,"max_input_tokens":1000000,"max_output_tokens":128000,"mode":"chat","supports_adaptive_thinking":true,"thinking_on_by_default":true,"supports_mid_conversation_system":true,"supports_assistant_prefill":false,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_sampling_params":false,"supports_tool_choice":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"supports_native_structured_output":false,"supports_max_reasoning_effort":true,"supports_output_config":true,"bedrock_output_config_effort_ceiling":"xhigh","supports_parallel_tool_use_config":true,"supports_forced_tool_use":true,"thinking_always_on":false,"prompt_cache_min_tokens":512,"source":"https://aws.amazon.com/bedrock/pricing/","provider":"bedrock","base_model":"claude-haiku-5-5","supports_audio_input":false,"supports_audio_output":false,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"],"max_tokens":128000},"databricks/system.ai.gemini-4-hydrogen":{"max_input_tokens":1048576,"max_output_tokens":65536,"metadata":{"notes":"Gemini 4 Hydrogen on Databricks is available only through Unity Gateway with the model service name system.ai.gemini-4-hydrogen on a global endpoint, and Databricks supports text input only. The Databricks pricing pages list no row for it yet, so input and output are the Global rates on the Vertex AI pricing page for Gemini 4 Hydrogen, no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},"mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models","supported_modalities":["text"],"supported_output_modalities":["text"],"supports_audio_input":false,"supports_reasoning":true,"supports_video_input":false,"supports_vision":false,"provider":"databricks","base_model":"system.ai.gemini-4-hydrogen"},"gpt-6.2-astra":{"fedramp_uplift_multiplier":1.1,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6.2-astra","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","base_model":"gpt-6.2-astra","web_search_billing_unit":"per_request","min_output_tokens":16,"service_tiers":["auto","default","priority","fast","ultrafast"],"supports_service_tier":true,"supports_custom_tools":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_native_structured_output":true,"supports_image_input":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_low_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"regional_endpoint_uplift_multiplier":1.1,"metadata":{"notes":"Context window 1,050,000 (922K input + 128K output). For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. 12-hour cache writes are priced for short-context requests only. Ultrafast is 3x Standard through 2027-02-28 and 6x Standard from 2027-03-01 (Pacific); *_ultrafast fields reflect 3x and must be updated on 2027-03-01. Knowledge cutoff: 2026-05-18. API rejects reasoning effort 'none'."}},"chatgpt/gpt-6.2-astra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6.2-astra"},"gpt-6.2-sol":{"fedramp_uplift_multiplier":1.1,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/models/gpt-6.2-sol","supported_endpoints":["/v1/chat/completions","/v1/responses","/v1/batch"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","base_model":"gpt-6.2-sol","web_search_billing_unit":"per_request","min_output_tokens":16,"service_tiers":["auto","default","priority","fast","ultrafast"],"supports_service_tier":true,"supports_custom_tools":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_native_structured_output":true,"supports_image_input":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"default_reasoning_effort":"medium","supports_low_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"regional_endpoint_uplift_multiplier":1.1,"metadata":{"notes":"Context window 1,050,000 (922K input + 128K output). For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. 12-hour cache writes are priced for short-context requests only. Ultrafast is 3x Standard through 2027-02-28 and 6x Standard from 2027-03-01 (Pacific); *_ultrafast fields reflect 3x and must be updated on 2027-03-01. Knowledge cutoff: 2026-06-01. API rejects reasoning effort 'none'."}},"chatgpt/gpt-6.2-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6.2-sol"},"gpt-6.2-astra-law":{"fedramp_uplift_multiplier":1.1,"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/batch","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"openai","base_model":"gpt-6.2-astra-law","web_search_billing_unit":"per_request"},"chatgpt/gpt-6.2-astra-law":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://developers.openai.com/api/docs/pricing","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_xhigh_reasoning_effort":true,"provider":"chatgpt","base_model":"gpt-6.2-astra-law"},"azure/gpt-6.2-astra":{"deprecation_date":"2027-02-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://ai.azure.com/catalog/models/gpt-6.2-astra","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-astra","min_output_tokens":16,"supports_service_tier":false,"supports_custom_tools":true,"supports_tool_search":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_native_structured_output":true,"supports_image_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_low_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"reasoning_contexts":["auto","current_turn","all_turns"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"web_search_billing_unit":"per_request","regional_processing_uplift_multiplier_us":1.1,"regional_processing_uplift_multiplier_eu":1.2,"max_tokens":128000,"metadata":{"notes":"PROVISIONAL: Azure has not yet published GPT-6.2 prices (catalog version 2026-11-17 links to https://aka.ms/AzureOAIpricing with no figures). Global Standard is set to OpenAI's Standard rates, as Azure matched OpenAI for GPT-6 Astra/Sol. Data Zone uplift follows the GPT-6 pattern: US 1.1x, EU 1.2x. Over 272K input tokens: 2x input/cache, 1.5x output for the full request. Knowledge cutoff: 2026-05-18. API rejects reasoning effort 'none'. Re-verify once Azure posts pricing."}},"azure/us/gpt-6.2-astra":{"deprecation_date":"2027-02-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-astra"},"azure/eu/gpt-6.2-astra":{"deprecation_date":"2027-02-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-astra"},"azure/apac/gpt-6.2-astra":{"deprecation_date":"2027-02-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"apac/gpt-6.2-astra"},"azure/gpt-6.2-sol":{"deprecation_date":"2027-11-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://ai.azure.com/catalog/models/gpt-6.2-sol","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-sol","min_output_tokens":16,"supports_service_tier":false,"supports_custom_tools":true,"supports_tool_search":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_native_structured_output":true,"supports_image_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_low_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"reasoning_contexts":["auto","current_turn","all_turns"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"web_search_billing_unit":"per_request","regional_processing_uplift_multiplier_us":1.1,"regional_processing_uplift_multiplier_eu":1.2,"max_tokens":128000,"metadata":{"notes":"PROVISIONAL: Azure has not yet published GPT-6.2 prices (catalog version 2026-11-17 links to https://aka.ms/AzureOAIpricing with no figures). Global Standard is set to OpenAI's Standard rates, as Azure matched OpenAI for GPT-6 Astra/Sol. Data Zone uplift follows the GPT-6 pattern: US 1.1x, EU 1.2x. Over 272K input tokens: 2x input/cache, 1.5x output for the full request. Knowledge cutoff: 2026-06-01. API rejects reasoning effort 'none'. Re-verify once Azure posts pricing."}},"azure/us/gpt-6.2-sol":{"deprecation_date":"2027-11-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-sol"},"azure/eu/gpt-6.2-sol":{"deprecation_date":"2027-11-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.2-sol"},"azure/apac/gpt-6.2-sol":{"deprecation_date":"2027-11-17","deprecation_source":"https://management.azure.com/subscriptions/{subscriptionId}/providers/Microsoft.CognitiveServices/locations/eastus/models?api-version=2024-10-01","max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/t5/microsoft-foundry-blog/gpt-6-2-astra-and-gpt-6-2-sol-now-generally-available-in/ba-p/4574122","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"apac/gpt-6.2-sol"},"azure/Microsoft-Decision-1.1":{"max_input_tokens":32768,"mode":"decisions","source":"https://prices.azure.com/api/retail/prices?$filter=contains(productName,'Decision')","provider":"azure","base_model":"microsoft-decision-1.1"},"fal_ai/kling/v4/text-to-image":{"metadata":{"comment":"fal reports the charge for each request in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. One image is 1.0 unit at the default 2K resolution, 1.4921 units at 4K and 2.4921 units at 8K ($0.157), measured 2026-11-17. The per-image prices are used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/kling/v4/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"kling/v4/text-to-image"},"fal_ai/kling/v4/image-to-image":{"metadata":{"comment":"fal reports the charge for each request in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. One image with one input image is 1.0476 units at the default 2K resolution, 1.5397 units at 4K and 2.5397 units at 8K ($0.16), and each further input image adds 0.0476 units ($0.003), measured 2026-11-17. The per-image prices assume one input image and are used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/kling/v4/image-to-image","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"kling/v4/image-to-image"},"fal_ai/kling/v4/long-image":{"metadata":{"comment":"fal reports the charge for each request in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. One long image is 2.3362 units at the default 1K resolution and 2.5362 units at 2K, whatever its height, measured 2026-11-17. The per-image prices are used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/kling/v4/long-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"kling/v4/long-image"},"fal_ai/heygen/avatar-6/identity-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/identity-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/identity-to-video"},"fal_ai/heygen/avatar-6/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/image-to-video"},"fal_ai/heygen/avatar-6/ref-video-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/ref-video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/ref-video-to-video"},"fal_ai/heygen/avatar-6/turbo/identity-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/turbo/identity-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/turbo/identity-to-video"},"fal_ai/heygen/avatar-6/turbo/image-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/turbo/image-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/turbo/image-to-video"},"fal_ai/heygen/avatar-6/turbo/ref-video-to-video":{"metadata":{"comment":"per second of generated video by resolution; output_cost_per_second is the default resolution 1080p rate"},"mode":"video_generation","source":"https://fal.ai/models/heygen/avatar-6/turbo/ref-video-to-video","supported_endpoints":["/v1/videos"],"supported_modalities":["text","video"],"supported_output_modalities":["video"],"provider":"fal_ai","base_model":"heygen/avatar-6/turbo/ref-video-to-video"},"fal_ai/microsoft/mai-image-2.7/text-to-image":{"metadata":{"comment":"fal bills this endpoint by tokens ($5 per million text input tokens, $8 per million image input tokens, $34 per million image output tokens) and reports the charge in dollars in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are the charges measured 2026-11-17 for one square image at the request's resolution field (1K default, 2K) with a short prompt, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/microsoft/mai-image-2.7/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"mai-image-2.7/text-to-image"},"fal_ai/microsoft/mai-image-2.7/edit":{"metadata":{"comment":"fal bills this endpoint by tokens ($5 per million text input tokens, $8 per million image input tokens, $34 per million image output tokens) and reports the charge in dollars in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are the charges measured 2026-11-17 for one square image at the request's resolution field (1K default, 2K) with a short prompt and one input image, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/microsoft/mai-image-2.7/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"mai-image-2.7/edit"},"fal_ai/microsoft/mai-image-2.7/flash/text-to-image":{"metadata":{"comment":"fal bills this endpoint by tokens ($1.75 per million text input tokens, $2.50 per million image input tokens, $10 per million image output tokens) and reports the charge in dollars in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are the charges measured 2026-11-17 for one square image at the request's resolution field (1K default, 2K) with a short prompt, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/microsoft/mai-image-2.7/flash/text-to-image","supported_endpoints":["/v1/images/generations"],"provider":"fal_ai","base_model":"mai-image-2.7/flash/text-to-image"},"fal_ai/microsoft/mai-image-2.7/flash/edit":{"metadata":{"comment":"fal bills this endpoint by tokens ($1.75 per million text input tokens, $2.50 per million image input tokens, $10 per million image output tokens) and reports the charge in dollars in the x-fal-billable-units response header, priced at output_cost_per_billable_unit dollars per unit. The per-image prices are the charges measured 2026-11-17 for one square image at the request's resolution field (1K default, 2K) with a short prompt and one input image, used when the header is absent"},"mode":"image_generation","source":"https://fal.ai/models/microsoft/mai-image-2.7/flash/edit","supported_endpoints":["/v1/images/edits","/v1/images/generations"],"supports_vision":true,"provider":"fal_ai","base_model":"mai-image-2.7/flash/edit"},"github_copilot/mai-code-2":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"responses","source":"https://docs.github.com/en/copilot/reference/copilot-billing/models-and-pricing","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"mai-code-2"},"github_copilot/mai-code-2-flash":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"responses","source":"https://docs.github.com/en/copilot/reference/copilot-billing/models-and-pricing","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"mai-code-2-flash"},"github_copilot/gpt-6.2-astra":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6.2-astra"},"github_copilot/gpt-6.2-sol":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"responses","source":"https://raw.githubusercontent.com/github/docs/main/data/tables/copilot/models-and-pricing.yml","supported_endpoints":["/v1/responses"],"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"provider":"github_copilot","base_model":"gpt-6.2-sol"},"openai/gpt-5.4-pro":{"max_tokens":1000000,"max_input_tokens":1000000,"max_output_tokens":100000,"provider":"openai","mode":"chat","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true,"supports_response_schema":true,"supports_system_messages":true,"supports_reasoning":true,"supports_web_search":true,"base_model":"gpt-5.4-pro","web_search_billing_unit":"per_request","supports_custom_tools":true},"replicate/google/imagen-4":{"mode":"image_generation","provider":"replicate","base_model":"imagen-4"},"replicate/black-forest-labs/flux-2-max":{"mode":"image_generation","provider":"replicate","base_model":"flux-2-max"},"replicate/google/nano-banana-pro":{"mode":"image_generation","provider":"replicate","base_model":"nano-banana-pro"},"replicate/black-forest-labs/flux-2-pro":{"mode":"image_generation","provider":"replicate","base_model":"flux-2-pro"},"replicate/google/imagen-4-ultra":{"mode":"image_generation","provider":"replicate","base_model":"imagen-4-ultra"},"replicate/google/imagen-4-fast":{"mode":"image_generation","provider":"replicate","base_model":"imagen-4-fast"},"replicate/google/imagen-3":{"mode":"image_generation","provider":"replicate","base_model":"imagen-3"},"replicate/google/imagen-3-fast":{"mode":"image_generation","provider":"replicate","base_model":"imagen-3-fast"},"replicate/openai/gpt-image-1.5":{"mode":"image_generation","provider":"replicate","base_model":"gpt-image-1.5"},"replicate/google/veo-3":{"mode":"video_generation","provider":"replicate","base_model":"veo-3"},"replicate/google/veo-3-fast":{"mode":"video_generation","provider":"replicate","base_model":"veo-3-fast"},"replicate/google/veo-3.1":{"mode":"video_generation","provider":"replicate","base_model":"veo-3.1"},"replicate/google/veo-3.1-fast":{"mode":"video_generation","provider":"replicate","base_model":"veo-3.1-fast"},"replicate/google/veo-2":{"mode":"video_generation","provider":"replicate","base_model":"veo-2"},"replicate/minimax/hailuo-02":{"mode":"video_generation","provider":"replicate","base_model":"hailuo-02"},"replicate/minimax/hailuo-02-fast":{"mode":"video_generation","provider":"replicate","base_model":"hailuo-02-fast"},"replicate/minimax/video-01":{"mode":"video_generation","provider":"replicate","base_model":"video-01"},"replicate/minimax/video-01-live":{"mode":"video_generation","provider":"replicate","base_model":"video-01-live"},"replicate/minimax/video-01-director":{"mode":"video_generation","provider":"replicate","base_model":"video-01-director"},"replicate/kwaivgi/kling-v2.1":{"mode":"video_generation","provider":"replicate","base_model":"kwaivgi/kling"},"replicate/kwaivgi/kling-v2.0":{"mode":"video_generation","provider":"replicate","base_model":"kwaivgi/kling"},"replicate/kwaivgi/kling-v1.6-pro":{"mode":"video_generation","provider":"replicate","base_model":"kwaivgi/kling-v1.6-pro"},"replicate/kwaivgi/kling-v1.6-standard":{"mode":"video_generation","provider":"replicate","base_model":"kwaivgi/kling-v1.6-standard"},"replicate/leonardoai/motion-2.0":{"mode":"video_generation","provider":"replicate","base_model":"leonardoai/motion-2.0"},"replicate/wan-video/wan-2.2-t2v-fast":{"mode":"video_generation","provider":"replicate","base_model":"wan-video/wan-2.2-t2v-fast"},"replicate/wan-video/wan-2.2-i2v-fast":{"mode":"video_generation","provider":"replicate","base_model":"wan-video/wan-2.2-i2v-fast"},"replicate/wan-video/wan-2.2-5b-fast":{"mode":"video_generation","provider":"replicate","base_model":"wan-video/wan-2.2-5b-fast"},"replicate/luma/ray-2-540p":{"mode":"video_generation","provider":"replicate","base_model":"luma/ray-2-540p"},"replicate/luma/ray-2-720p":{"mode":"video_generation","provider":"replicate","base_model":"luma/ray-2-720p"},"replicate/luma/ray-flash-2-540p":{"mode":"video_generation","provider":"replicate","base_model":"luma/ray-flash-2-540p"},"replicate/luma/ray-flash-2-720p":{"mode":"video_generation","provider":"replicate","base_model":"luma/ray-flash-2-720p"},"replicate/pixverse/pixverse-v4.5":{"mode":"video_generation","provider":"replicate","base_model":"pixverse/pixverse"},"replicate/pixverse/pixverse-v4":{"mode":"video_generation","provider":"replicate","base_model":"pixverse/pixverse-v4"},"bedrock/amazon.nova-canvas-v1:0":{"deprecation_date":"2026-09-30","deprecation_source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-lifecycle-legacy.html","max_input_tokens":2600,"mode":"image_generation","supports_nova_canvas_image_edit":true,"provider":"bedrock","base_model":"nova-canvas","is_deprecated":true},"stability.stable-diffusion-xl-v1":{"max_input_tokens":77,"mode":"image_generation","provider":"bedrock","base_model":"stable-diffusion-xl-v1"},"stability.stable-diffusion-xl-v0":{"max_input_tokens":77,"mode":"image_generation","provider":"bedrock","base_model":"stable-diffusion-xl-v0"},"azure/dall-e-3":{"mode":"image_generation","source":"https://azure.microsoft.com/en-us/pricing/details/azure-openai/","provider":"azure","base_model":"dall-e-3"},"openrouter/heygen/heygen-avatar-6-turbo":{"provider":"openrouter","base_model":"heygen-avatar-6-turbo","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/heygen/heygen-avatar-6":{"provider":"openrouter","base_model":"heygen-avatar-6","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/openai/gpt-6.2-astra-pro":{"provider":"openrouter","base_model":"gpt-6.2-astra-pro","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-astra-pro:batch":{"provider":"openrouter","base_model":"gpt-6.2-astra-pro:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-astra":{"provider":"openrouter","base_model":"gpt-6.2-astra","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-astra:batch":{"provider":"openrouter","base_model":"gpt-6.2-astra:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-sol-pro":{"provider":"openrouter","base_model":"gpt-6.2-sol-pro","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-sol-pro:batch":{"provider":"openrouter","base_model":"gpt-6.2-sol-pro:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-sol":{"provider":"openrouter","base_model":"gpt-6.2-sol","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.2-sol:batch":{"provider":"openrouter","base_model":"gpt-6.2-sol:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/microsoft/mai-image-2.7-flash":{"provider":"openrouter","base_model":"mai-image-2.7-flash","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/microsoft/mai-image-2.7":{"provider":"openrouter","base_model":"mai-image-2.7","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/microsoft/microsoft-decision-1.1":{"provider":"openrouter","base_model":"microsoft-decision-1.1","mode":"decisions","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_vision":true,"supports_decisions":true},"openrouter/bytedance/seedance-2.5-flash":{"provider":"openrouter","base_model":"seedance-2.5-flash","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/liquid/d1.2-ultra":{"provider":"openrouter","base_model":"d1.2-ultra","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true},"openrouter/liquid/d1.2":{"provider":"openrouter","base_model":"d1.2","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true},"openrouter/tencent/hy4-flash":{"provider":"openrouter","base_model":"hy4-flash","mode":"chat","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/qwen/qwen4-max":{"provider":"openrouter","base_model":"qwen4-max","mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/recraft/recraft-v4.1-flash-v2":{"provider":"openrouter","base_model":"recraft-v4.1-flash-v2","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982},"openrouter/recraft/recraft-v4.1-flash-v2:free":{"provider":"openrouter","base_model":"recraft-v4.1-flash-v2:free","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982},"openrouter/tencent/hy-world-2.1-panorama":{"provider":"openrouter","base_model":"hy-world-2.1-panorama","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686},"openrouter/inclusionai/ling-3.2-flash:free":{"provider":"openrouter","base_model":"ling-3.2-flash:free","mode":"chat","max_input_tokens":1048576,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/heygen/heygen-decision-1-turbo":{"provider":"openrouter","base_model":"heygen-decision-1-turbo","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/heygen/heygen-decision-1":{"provider":"openrouter","base_model":"heygen-decision-1","mode":"decisions","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_vision":true,"supports_decisions":true},"openrouter/bytedance-seed/seedream-5-5":{"provider":"openrouter","base_model":"seedream-5-5","mode":"image_generation","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/perplexity/pplx-decider-flash-v1.3:free":{"provider":"openrouter","base_model":"pplx-decider-flash-v1.3:free","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/perplexity/pplx-decider-v1.3:free":{"provider":"openrouter","base_model":"pplx-decider-v1.3:free","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/google/gemini-4-hydrogen":{"provider":"openrouter","base_model":"gemini-4-hydrogen","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/google/gemini-4-hydrogen:batch":{"provider":"openrouter","base_model":"gemini-4-hydrogen:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/speechify/simba-3.5":{"provider":"openrouter","base_model":"simba-3.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/google/gemini-decision-1":{"provider":"openrouter","base_model":"gemini-decision-1","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_decisions":true},"openrouter/minimax/minimax-h3-image-highspeed":{"provider":"openrouter","base_model":"minimax-h3-image-highspeed","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"supports_vision":true},"openrouter/minimax/minimax-h3-image":{"provider":"openrouter","base_model":"minimax-h3-image","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"supports_vision":true},"openrouter/baidu/ernie-5.5-flash-preview:free":{"provider":"openrouter","base_model":"ernie-5.5-flash-preview:free","mode":"chat","max_input_tokens":1024000,"max_output_tokens":921600,"max_tokens":921600,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true},"openrouter/baidu/ernie-light-dt-1.0-intl":{"provider":"openrouter","base_model":"ernie-light-dt-1.0-intl","mode":"chat","max_input_tokens":1024000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/x-ai/grok-stt-2.0":{"provider":"openrouter","base_model":"grok-stt-2.0","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/google/gemini-4-carbon":{"provider":"openrouter","base_model":"gemini-4-carbon","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/google/gemini-4-carbon:batch":{"provider":"openrouter","base_model":"gemini-4-carbon:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/nex-agi/nex-n2.6":{"provider":"openrouter","base_model":"nex-n2.6","mode":"chat","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/nace-ai/drex-max-finance":{"provider":"openrouter","base_model":"drex-max-finance","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/thinkingmachines/inkling-2-small":{"provider":"openrouter","base_model":"inkling-2-small","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/thinkingmachines/inkling-2-small:free":{"provider":"openrouter","base_model":"inkling-2-small:free","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_reasoning":true,"supports_vision":true},"openrouter/anthropic/claude-fable-5.2":{"provider":"openrouter","base_model":"claude-fable-5.2","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/anthropic/claude-fable-5.2:batch":{"provider":"openrouter","base_model":"claude-fable-5.2:batch","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/meta/muse-ember":{"provider":"openrouter","base_model":"muse-ember","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/liquid/lfm-3-1.7b":{"provider":"openrouter","base_model":"lfm-3-1.7b","mode":"chat","max_input_tokens":262144,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/liquid/lfm-3-1.7b:free":{"provider":"openrouter","base_model":"lfm-3-1.7b:free","mode":"chat","max_input_tokens":262144,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true},"openrouter/parallel/decisions-fast":{"provider":"openrouter","base_model":"decisions-fast","mode":"decisions","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"supports_decisions":true},"openrouter/google/gemini-omni-1.2-flash":{"provider":"openrouter","base_model":"gemini-omni-1.2-flash","mode":"chat","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"supports_reasoning":true,"supports_vision":true},"openrouter/kt/midm-k-3.0-pro":{"provider":"openrouter","base_model":"midm-k-3.0-pro","mode":"chat","max_input_tokens":261795,"max_output_tokens":235615,"max_tokens":235615,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/openai/gpt-image-2.6-sunburst":{"provider":"openrouter","base_model":"gpt-image-2.6-sunburst","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/openai/gpt-image-2.6-flare":{"provider":"openrouter","base_model":"gpt-image-2.6-flare","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/google/gemini-omni-1.1-flash":{"provider":"openrouter","base_model":"gemini-omni-1.1-flash","mode":"chat","max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"supports_reasoning":true,"supports_vision":true},"openrouter/ricursiveai/ripple-65b-preview":{"provider":"openrouter","base_model":"ripple-65b-preview","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/tencent/hy4":{"provider":"openrouter","base_model":"hy4","mode":"chat","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/morph/reflex":{"provider":"openrouter","base_model":"reflex","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/cohere/embed-5-fast":{"provider":"openrouter","base_model":"embed-5-fast","mode":"embedding","max_input_tokens":128000,"max_output_tokens":115200,"max_tokens":115200,"supports_vision":true},"openrouter/cohere/embed-5-pro":{"provider":"openrouter","base_model":"embed-5-pro","mode":"embedding","max_input_tokens":128000,"max_output_tokens":115200,"max_tokens":115200,"supports_vision":true},"openrouter/cohere/rerank-5-fast":{"provider":"openrouter","base_model":"rerank-5-fast","mode":"chat","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491},"openrouter/cohere/rerank-5-pro":{"provider":"openrouter","base_model":"rerank-5-pro","mode":"chat","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491},"openrouter/cloudflare/clairic":{"provider":"openrouter","base_model":"clairic","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true},"openrouter/moondream/parakeet-ultra":{"provider":"openrouter","base_model":"parakeet-ultra","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/motif-technologies/motif-3":{"provider":"openrouter","base_model":"motif-3","mode":"chat","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/z-ai/glm-5.3-flashx-legal":{"provider":"openrouter","base_model":"glm-5.3-flashx-legal","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/tbc/mischief-1":{"provider":"openrouter","base_model":"mischief-1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/baseten/joey-1":{"provider":"openrouter","base_model":"joey-1","mode":"decisions","max_input_tokens":16384,"max_output_tokens":14745,"max_tokens":14745,"supports_vision":true,"supports_decisions":true},"openrouter/sourceful/riverflow-v2.6-pro":{"provider":"openrouter","base_model":"riverflow-v2.6-pro","mode":"image_generation","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_reasoning":true,"supports_vision":true},"openrouter/nvidia/nemotron-3.5-transcribe":{"provider":"openrouter","base_model":"nemotron-3.5-transcribe","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/assemblyai/assemblyai-tts-1":{"provider":"openrouter","base_model":"assemblyai-tts-1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/particle-ai/particle-1":{"provider":"openrouter","base_model":"particle-1","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/particle-ai/particle-1:free":{"provider":"openrouter","base_model":"particle-1:free","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/liquid/d1.1":{"provider":"openrouter","base_model":"d1.1","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/~liquid/d1-ultra-latest":{"provider":"openrouter","base_model":"d1-ultra-latest","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true},"openrouter/~liquid/d1-latest":{"provider":"openrouter","base_model":"d1-latest","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true},"openrouter/amazon/titan-embed-image-v1":{"provider":"openrouter","base_model":"titan-embed-image-v1","mode":"embedding","max_input_tokens":128,"max_output_tokens":115,"max_tokens":115,"supports_vision":true},"openrouter/amazon/titan-embed-text-v1":{"provider":"openrouter","base_model":"titan-embed-text-v1","mode":"embedding","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372},"openrouter/amazon/titan-embed-text-v2":{"provider":"openrouter","base_model":"titan-embed-text-v2","mode":"embedding","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372},"openrouter/openai/gpt-6.1-terra-pro":{"provider":"openrouter","base_model":"gpt-6.1-terra-pro","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.1-terra-pro:batch":{"provider":"openrouter","base_model":"gpt-6.1-terra-pro:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.1-terra":{"provider":"openrouter","base_model":"gpt-6.1-terra","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/openai/gpt-6.1-terra:batch":{"provider":"openrouter","base_model":"gpt-6.1-terra:batch","mode":"chat","max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/heygen/heygen-voice-1-turbo":{"provider":"openrouter","base_model":"heygen-voice-1-turbo","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/heygen/heygen-voice-1-turbo:free":{"provider":"openrouter","base_model":"heygen-voice-1-turbo:free","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/moonshotai/kimi-k3.1-pro":{"provider":"openrouter","base_model":"kimi-k3.1-pro","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/moonshotai/kimi-k3.1-pro:batch":{"provider":"openrouter","base_model":"kimi-k3.1-pro:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_reasoning":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/recursive/flint-v1.0:free":{"provider":"openrouter","base_model":"flint-v1.0:free","mode":"chat","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"openrouter/baseten/mammoth-1":{"provider":"openrouter","base_model":"mammoth-1","mode":"decisions","max_input_tokens":16384,"max_output_tokens":14745,"max_tokens":14745,"supports_vision":true,"supports_decisions":true},"openrouter/baseten/elephant-1":{"provider":"openrouter","base_model":"elephant-1","mode":"decisions","max_input_tokens":16384,"max_output_tokens":14745,"max_tokens":14745,"supports_vision":true,"supports_decisions":true},"openrouter/nace-ai/drex-max":{"provider":"openrouter","base_model":"drex-max","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/nvidia/nemotron-3.5-embed-1b:free":{"provider":"openrouter","base_model":"nemotron-3.5-embed-1b:free","mode":"embedding","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_vision":true},"openrouter/nvidia/nemotron-3.5-embed-8b:free":{"provider":"openrouter","base_model":"nemotron-3.5-embed-8b:free","mode":"embedding","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_vision":true},"openrouter/nvidia/nemotron-3.5-rerank-8b:free":{"provider":"openrouter","base_model":"nemotron-3.5-rerank-8b:free","mode":"chat","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_vision":true},"openrouter/mistralai/voxtral-mini-tts-3-0":{"provider":"openrouter","base_model":"voxtral-mini-tts-3-0","mode":"chat","max_input_tokens":4096,"max_output_tokens":3276,"max_tokens":3276},"openrouter/mistralai/voxtral-small-transcribe-3-0":{"provider":"openrouter","base_model":"voxtral-small-transcribe-3-0","mode":"chat","max_input_tokens":65536,"max_output_tokens":52428,"max_tokens":52428,"supports_audio_input":true},"openrouter/mistralai/voxtral-mini-transcribe-3-0":{"provider":"openrouter","base_model":"voxtral-mini-transcribe-3-0","mode":"chat","max_input_tokens":65536,"max_output_tokens":52428,"max_tokens":52428,"supports_audio_input":true},"openrouter/minimax/minimax-m3.1-flash":{"provider":"openrouter","base_model":"minimax-m3.1-flash","mode":"chat","max_input_tokens":1048576,"max_output_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/deepseek/deepseek-v4.1-flash-1017":{"provider":"openrouter","base_model":"deepseek-v4.1-flash-1017","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/deepseek/deepseek-v4.1-flash-1017:batch":{"provider":"openrouter","base_model":"deepseek-v4.1-flash-1017:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/elorian/glimpse":{"provider":"openrouter","base_model":"glimpse","mode":"chat","max_input_tokens":262144,"max_output_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/qwen/qwen-image-2.1-turbo":{"provider":"openrouter","base_model":"qwen-image-2.1-turbo","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_response_schema":true,"supports_vision":true},"openrouter/qwen/qwen-image-2.1-pro":{"provider":"openrouter","base_model":"qwen-image-2.1-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_response_schema":true,"supports_vision":true},"openrouter/perplexity/pplx-decider-flash-v1.2":{"provider":"openrouter","base_model":"pplx-decider-flash-v1.2","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/perplexity/pplx-decider-v1.2":{"provider":"openrouter","base_model":"pplx-decider-v1.2","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/~perplexity/pplx-decider-flash-latest":{"provider":"openrouter","base_model":"pplx-decider-flash-latest","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/reflection/beam-1.0-501b-a23b":{"provider":"openrouter","base_model":"beam-1.0-501b-a23b","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/prism-ml/ternary-bonsai-2-9b":{"provider":"openrouter","base_model":"ternary-bonsai-2-9b","mode":"chat","max_input_tokens":262144,"max_output_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/respan/span-02":{"provider":"openrouter","base_model":"span-02","mode":"decisions","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_vision":true,"supports_decisions":true},"openrouter/liquid/d1-ultra":{"provider":"openrouter","base_model":"d1-ultra","mode":"decisions","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/krea/k3-edit":{"provider":"openrouter","base_model":"k3-edit","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/moondream/moondream3-preview":{"provider":"openrouter","base_model":"moondream3-preview","mode":"chat","max_input_tokens":32000,"max_output_tokens":4096,"max_tokens":4096,"supports_reasoning":true,"supports_vision":true},"openrouter/moondream/moondream3.1-9b-a2b":{"provider":"openrouter","base_model":"moondream3.1-9b-a2b","mode":"chat","max_input_tokens":32000,"max_output_tokens":4096,"max_tokens":4096,"supports_reasoning":true,"supports_vision":true},"openrouter/meta/muse-spark-open":{"provider":"openrouter","base_model":"muse-spark-open","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/black-forest-labs/flux-3-image-klein":{"provider":"openrouter","base_model":"flux-3-image-klein","mode":"image_generation","max_input_tokens":46864,"max_output_tokens":42177,"max_tokens":42177,"supports_vision":true},"openrouter/nvidia/nemotron-3.5-super-vl":{"provider":"openrouter","base_model":"nemotron-3.5-super-vl","mode":"chat","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/nvidia/nemotron-3.5-super-vl:free":{"provider":"openrouter","base_model":"nemotron-3.5-super-vl:free","mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/unbiased/pareto-26.10":{"provider":"openrouter","base_model":"pareto-26.10","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/typesafe/jev-1.14":{"provider":"openrouter","base_model":"jev-1.14","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/assemblyai/universal-3-6-pro":{"provider":"openrouter","base_model":"universal-3-6-pro","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/~perplexity/pplx-decider-latest":{"provider":"openrouter","base_model":"pplx-decider-latest","mode":"decisions","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_vision":true,"supports_decisions":true},"openrouter/~unbiased/pareto-latest":{"provider":"openrouter","base_model":"pareto-latest","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/unbiased/pareto-26.9":{"provider":"openrouter","base_model":"pareto-26.9","mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/heygen/heygen-voice-1":{"provider":"openrouter","base_model":"heygen-voice-1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/liquid/lfm-2.5-24b-a2b":{"provider":"openrouter","base_model":"lfm-2.5-24b-a2b","mode":"chat","max_input_tokens":128000,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/fish-audio/drama-3":{"provider":"openrouter","base_model":"drama-3","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/google/gemini-4-argon":{"provider":"openrouter","base_model":"gemini-4-argon","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/google/gemini-4-argon:batch":{"provider":"openrouter","base_model":"gemini-4-argon:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/scaledown/summarize":{"provider":"openrouter","base_model":"summarize","mode":"chat","max_input_tokens":1000000,"max_output_tokens":4000,"max_tokens":4000,"supports_vision":true,"supports_pdf_input":true},"openrouter/scaledown/extract":{"provider":"openrouter","base_model":"extract","mode":"chat","max_input_tokens":1000000,"max_output_tokens":900000,"max_tokens":900000,"supports_vision":true,"supports_pdf_input":true},"openrouter/scaledown/classify-1":{"provider":"openrouter","base_model":"classify-1","mode":"decisions","max_input_tokens":1000000,"max_output_tokens":900000,"max_tokens":900000,"supports_vision":true,"supports_pdf_input":true,"supports_decisions":true},"openrouter/nace-ai/drex-v2.0":{"provider":"openrouter","base_model":"drex-v2.0","mode":"decisions","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_decisions":true},"openrouter/meta/muse-wonder":{"provider":"openrouter","base_model":"muse-wonder","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/microsoft/microsoft-decision-1":{"provider":"openrouter","base_model":"microsoft-decision-1","mode":"decisions","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_decisions":true},"openrouter/nace-ai/drex-v1.5":{"provider":"openrouter","base_model":"drex-v1.5","mode":"decisions","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_decisions":true},"openrouter/cloudflare/clef-omni":{"provider":"openrouter","base_model":"clef-omni","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_audio_input":true,"supports_decisions":true},"openrouter/stepfun/step-5-preview":{"provider":"openrouter","base_model":"step-5-preview","mode":"chat","max_input_tokens":1000000,"max_output_tokens":64000,"max_tokens":64000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/upstage/solar-decide-flash":{"provider":"openrouter","base_model":"solar-decide-flash","mode":"decisions","max_input_tokens":524288,"max_output_tokens":471859,"max_tokens":471859,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/anthropic/claude-haiku-5.5":{"provider":"openrouter","base_model":"claude-haiku-5.5","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/anthropic/claude-haiku-5.5:batch":{"provider":"openrouter","base_model":"claude-haiku-5.5:batch","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/elevenlabs/eleven-v4":{"provider":"openrouter","base_model":"eleven-v4","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-v4-turbo":{"provider":"openrouter","base_model":"eleven-v4-turbo","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-v3":{"provider":"openrouter","base_model":"eleven-v3","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-v3-conversational":{"provider":"openrouter","base_model":"eleven-v3-conversational","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-flash-v2.5":{"provider":"openrouter","base_model":"eleven-flash-v2.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-turbo-v2":{"provider":"openrouter","base_model":"eleven-turbo-v2","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-multilingual-v2":{"provider":"openrouter","base_model":"eleven-multilingual-v2","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-turbo-v2.5":{"provider":"openrouter","base_model":"eleven-turbo-v2.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/eleven-flash-v2":{"provider":"openrouter","base_model":"eleven-flash-v2","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/elevenlabs/scribe-v2":{"provider":"openrouter","base_model":"scribe-v2","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/elevenlabs/scribe-v2-medical":{"provider":"openrouter","base_model":"scribe-v2-medical","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/openai/gpt-6-luna-decisions":{"provider":"openrouter","base_model":"gpt-6-luna-decisions","mode":"decisions","max_input_tokens":1050000,"max_output_tokens":945000,"max_tokens":945000,"supports_vision":true,"supports_decisions":true},"openrouter/x-ai/grok-imagine-video-1.5-lite":{"provider":"openrouter","base_model":"grok-imagine-video-1.5-lite","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/google/gemini-nano-banana-2.1":{"provider":"openrouter","base_model":"gemini-nano-banana-2.1","mode":"chat","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/mistralai/mistral-large-4-0":{"provider":"openrouter","base_model":"mistral-large-4-0","mode":"chat","max_input_tokens":1048576,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/tencent/hy-image-v3.5-preview":{"provider":"openrouter","base_model":"hy-image-v3.5-preview","mode":"image_generation","max_input_tokens":100000,"max_output_tokens":90000,"max_tokens":90000,"supports_vision":true},"openrouter/inclusionai/ling-3.1-flash":{"provider":"openrouter","base_model":"ling-3.1-flash","mode":"chat","max_input_tokens":1000000,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_prompt_caching":true},"openrouter/bytedance-seed/seedream-5-0-flash":{"provider":"openrouter","base_model":"seedream-5-0-flash","mode":"image_generation","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/black-forest-labs/flux-3-image":{"provider":"openrouter","base_model":"flux-3-image","mode":"image_generation","max_input_tokens":46864,"max_output_tokens":42177,"max_tokens":42177,"supports_vision":true},"openrouter/liquid/d1":{"provider":"openrouter","base_model":"d1","mode":"decisions","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/cloudflare/clef-flash":{"provider":"openrouter","base_model":"clef-flash","mode":"decisions","max_input_tokens":24576,"max_output_tokens":22118,"max_tokens":22118,"supports_vision":true,"supports_decisions":true},"openrouter/cloudflare/clef":{"provider":"openrouter","base_model":"clef","mode":"decisions","max_input_tokens":65537,"max_output_tokens":58983,"max_tokens":58983,"supports_vision":true,"supports_decisions":true},"openrouter/microsoft/mai-voice-2.1-flash":{"provider":"openrouter","base_model":"mai-voice-2.1-flash","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/microsoft/mai-voice-2.1":{"provider":"openrouter","base_model":"mai-voice-2.1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/unbiased/pareto-26.10-preview":{"provider":"openrouter","base_model":"pareto-26.10-preview","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/heygen/heygen-video-1":{"provider":"openrouter","base_model":"heygen-video-1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/togethercomputer/tev1-4b-experimental":{"provider":"openrouter","base_model":"tev1-4b-experimental","mode":"decisions","max_input_tokens":32768,"max_output_tokens":8,"max_tokens":8,"supports_decisions":true},"openrouter/mongodb/rerank-3-lite":{"provider":"openrouter","base_model":"rerank-3-lite","mode":"chat","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/mongodb/rerank-3":{"provider":"openrouter","base_model":"rerank-3","mode":"chat","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/anthropic/claude-sonnet-5.5":{"provider":"openrouter","base_model":"claude-sonnet-5.5","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/anthropic/claude-sonnet-5.5:batch":{"provider":"openrouter","base_model":"claude-sonnet-5.5:batch","mode":"chat","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/upstage/solar-decide":{"provider":"openrouter","base_model":"solar-decide","mode":"decisions","max_input_tokens":524288,"max_output_tokens":471859,"max_tokens":471859,"supports_decisions":true,"supports_prompt_caching":true},"openrouter/respan/span-01":{"provider":"openrouter","base_model":"span-01","mode":"decisions","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_decisions":true},"openrouter/respan/span-01-lite":{"provider":"openrouter","base_model":"span-01-lite","mode":"decisions","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_decisions":true},"openrouter/respan/span-01-lite:free":{"provider":"openrouter","base_model":"span-01-lite:free","mode":"decisions","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_decisions":true},"openrouter/bytedance-seed/seed-audio-1-0":{"provider":"openrouter","base_model":"seed-audio-1-0","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/jaredpalmer/kev-4b":{"provider":"openrouter","base_model":"kev-4b","mode":"decisions","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_decisions":true},"openrouter/perceptron/perceptron-mk1.5":{"provider":"openrouter","base_model":"perceptron-mk1.5","mode":"chat","max_input_tokens":36864,"max_output_tokens":8192,"max_tokens":8192,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true},"openrouter/google/gemini-3.5-transcribe":{"provider":"openrouter","base_model":"gemini-3.5-transcribe","mode":"chat","max_input_tokens":98304,"max_output_tokens":32768,"max_tokens":32768,"supports_audio_input":true},"openrouter/fish-audio/transcribe-1-pro":{"provider":"openrouter","base_model":"transcribe-1-pro","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_audio_input":true},"openrouter/fireworks/ember-1":{"provider":"openrouter","base_model":"ember-1","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/google/gemini-3.8-flash-lite-tts":{"provider":"openrouter","base_model":"gemini-3.8-flash-lite-tts","mode":"chat","max_input_tokens":32768,"max_output_tokens":16384,"max_tokens":16384},"openrouter/google/gemini-3.8-flash-tts":{"provider":"openrouter","base_model":"gemini-3.8-flash-tts","mode":"chat","max_input_tokens":32768,"max_output_tokens":16384,"max_tokens":16384},"openrouter/z-ai/glm-5.3-prime":{"provider":"openrouter","base_model":"glm-5.3-prime","mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/qwen/qwen3.8-max-prime":{"provider":"openrouter","base_model":"qwen3.8-max-prime","mode":"chat","max_input_tokens":1000000,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/recraft/recraft-v4.1-flash":{"provider":"openrouter","base_model":"recraft-v4.1-flash","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982},"openrouter/aion-labs/aion-3.5-mini":{"provider":"openrouter","base_model":"aion-3.5-mini","mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/aion-labs/aion-3.5":{"provider":"openrouter","base_model":"aion-3.5","mode":"chat","max_input_tokens":262144,"max_output_tokens":32768,"max_tokens":32768,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/upstage/solar-mini4":{"provider":"openrouter","base_model":"solar-mini4","mode":"chat","max_input_tokens":524288,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/x-ai/grok-4.7:batch":{"provider":"openrouter","base_model":"grok-4.7:batch","mode":"chat","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/black-forest-labs/flux-video-edit":{"provider":"openrouter","base_model":"flux-video-edit","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/openai/gpt-image-2.5-sunburst":{"provider":"openrouter","base_model":"gpt-image-2.5-sunburst","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/openai/gpt-image-2.5-flare":{"provider":"openrouter","base_model":"gpt-image-2.5-flare","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/microsoft/mai-image-2.6":{"provider":"openrouter","base_model":"mai-image-2.6","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/microsoft/mai-image-2.6-flash":{"provider":"openrouter","base_model":"mai-image-2.6-flash","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/qwen/qwen3.8-max-0902:batch":{"provider":"openrouter","base_model":"qwen3.8-max-0902:batch","mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/minimax/hailuo-3-max":{"provider":"openrouter","base_model":"hailuo-3-max","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/alibaba/wan-3.0-prime":{"provider":"openrouter","base_model":"wan-3.0-prime","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true},"openrouter/meta/muse-image":{"provider":"openrouter","base_model":"muse-image","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-styles-pro":{"provider":"openrouter","base_model":"recraft-v4-styles-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-styles-vector":{"provider":"openrouter","base_model":"recraft-v4-styles-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-styles-pro-vector":{"provider":"openrouter","base_model":"recraft-v4-styles-pro-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-styles":{"provider":"openrouter","base_model":"recraft-v4-styles","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/qwen/qwen3-reranker-0.6b":{"provider":"openrouter","base_model":"qwen3-reranker-0.6b","mode":"chat","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0},"openrouter/alibaba/wan-3.0":{"provider":"openrouter","base_model":"wan-3.0","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true,"supports_audio_input":true},"openrouter/heygen/avatar-iv":{"provider":"openrouter","base_model":"avatar-iv","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/black-forest-labs/flux-video-upscale":{"provider":"openrouter","base_model":"flux-video-upscale","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/qwen/qwen3.8-27b:batch":{"provider":"openrouter","base_model":"qwen3.8-27b:batch","mode":"chat","max_input_tokens":262144,"max_output_tokens":235929,"max_tokens":235929,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/bytedance-seed/seedream-5-0-lite":{"provider":"openrouter","base_model":"seedream-5-0-lite","mode":"image_generation","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/mongodb/voyage-code-4":{"provider":"openrouter","base_model":"voyage-code-4","mode":"embedding","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/qwen/qwen3-reranker-8b":{"provider":"openrouter","base_model":"qwen3-reranker-8b","mode":"chat","max_input_tokens":32768,"max_output_tokens":0,"max_tokens":0},"openrouter/bytedance-seed/seedream-5-0-pro":{"provider":"openrouter","base_model":"seedream-5-0-pro","mode":"image_generation","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/deepgram/flux-tts":{"provider":"openrouter","base_model":"flux-tts","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/bytedance/seedance-2.0-mini":{"provider":"openrouter","base_model":"seedance-2.0-mini","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/x-ai/grok-4.6:batch":{"provider":"openrouter","base_model":"grok-4.6:batch","mode":"chat","max_input_tokens":500000,"max_output_tokens":450000,"max_tokens":450000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_pdf_input":true,"supports_prompt_caching":true},"openrouter/x-ai/grok-imagine-image-2.0":{"provider":"openrouter","base_model":"grok-imagine-image-2.0","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_response_schema":true,"supports_vision":true},"openrouter/meta/muse-glimmer-30b:batch":{"provider":"openrouter","base_model":"muse-glimmer-30b:batch","mode":"chat","max_input_tokens":131072,"max_output_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/bytedance/seedance-2.5":{"provider":"openrouter","base_model":"seedance-2.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/inclusionai/ling-3.0-tiny":{"provider":"openrouter","base_model":"ling-3.0-tiny","mode":"chat","max_input_tokens":262144,"max_output_tokens":250000,"max_tokens":250000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/qwen/qwen-image-3":{"provider":"openrouter","base_model":"qwen-image-3","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_response_schema":true,"supports_vision":true},"openrouter/qwen/qwen-image-3-pro":{"provider":"openrouter","base_model":"qwen-image-3-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_response_schema":true,"supports_vision":true},"openrouter/black-forest-labs/flux-3-video":{"provider":"openrouter","base_model":"flux-3-video","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/deepseek/deepseek-v4-flash-0731:batch":{"provider":"openrouter","base_model":"deepseek-v4-flash-0731:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":384000,"max_tokens":384000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/thinkingmachines/inkling-small:batch":{"provider":"openrouter","base_model":"inkling-small:batch","mode":"chat","max_input_tokens":524288,"max_output_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_vision":true,"supports_audio_input":true,"supports_prompt_caching":true},"openrouter/minimax/hailuo-3":{"provider":"openrouter","base_model":"hailuo-3","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/fish-audio/s2.1-pro-free:free":{"provider":"openrouter","base_model":"s2.1-pro-free:free","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0},"openrouter/runway/aleph-2":{"provider":"openrouter","base_model":"aleph-2","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/runway/gen-4.5":{"provider":"openrouter","base_model":"gen-4.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/mongodb/rerank-2.5-lite":{"provider":"openrouter","base_model":"rerank-2.5-lite","mode":"chat","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/mongodb/rerank-2.5":{"provider":"openrouter","base_model":"rerank-2.5","mode":"chat","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/mongodb/voyage-multimodal-3.5":{"provider":"openrouter","base_model":"voyage-multimodal-3.5","mode":"embedding","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800,"supports_vision":true},"openrouter/mongodb/voyage-4-lite":{"provider":"openrouter","base_model":"voyage-4-lite","mode":"embedding","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/mongodb/voyage-4":{"provider":"openrouter","base_model":"voyage-4","mode":"embedding","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/mongodb/voyage-4-large":{"provider":"openrouter","base_model":"voyage-4-large","mode":"embedding","max_input_tokens":32000,"max_output_tokens":28800,"max_tokens":28800},"openrouter/microsoft/mai-image-2.5-pro":{"provider":"openrouter","base_model":"mai-image-2.5-pro","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/krea/krea-2-large":{"provider":"openrouter","base_model":"krea-2-large","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/krea/krea-2-medium":{"provider":"openrouter","base_model":"krea-2-medium","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/krea/krea-2-medium-turbo":{"provider":"openrouter","base_model":"krea-2-medium-turbo","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/x-ai/grok-imagine-video-1.5":{"provider":"openrouter","base_model":"grok-imagine-video-1.5","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/nvidia/nemotron-3-embed-1b:free":{"provider":"openrouter","base_model":"nemotron-3-embed-1b:free","mode":"embedding","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686},"openrouter/alibaba/happyhorse-1.1":{"provider":"openrouter","base_model":"happyhorse-1.1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/openai/gpt-image-2":{"provider":"openrouter","base_model":"gpt-image-2","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/openai/gpt-image-1":{"provider":"openrouter","base_model":"gpt-image-1","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/openai/gpt-image-1-mini":{"provider":"openrouter","base_model":"gpt-image-1-mini","mode":"image_generation","max_input_tokens":400000,"max_output_tokens":360000,"max_tokens":360000,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/alibaba/happyhorse-1.0":{"provider":"openrouter","base_model":"happyhorse-1.0","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/z-ai/glm-5.2:batch":{"provider":"openrouter","base_model":"glm-5.2:batch","mode":"chat","max_input_tokens":1048576,"max_output_tokens":943718,"max_tokens":943718,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_prompt_caching":true},"openrouter/sourceful/riverflow-v2.5-pro":{"provider":"openrouter","base_model":"riverflow-v2.5-pro","mode":"image_generation","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_reasoning":true,"supports_vision":true},"openrouter/sourceful/riverflow-v2.5-fast":{"provider":"openrouter","base_model":"riverflow-v2.5-fast","mode":"image_generation","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491,"supports_reasoning":true,"supports_vision":true},"openrouter/qwen/qwen3.7-plus:batch":{"provider":"openrouter","base_model":"qwen3.7-plus:batch","mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true},"openrouter/microsoft/mai-image-2.5":{"provider":"openrouter","base_model":"mai-image-2.5","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":1024,"max_tokens":1024,"supports_vision":true},"openrouter/minimax/minimax-m3:batch":{"provider":"openrouter","base_model":"minimax-m3:batch","mode":"chat","max_input_tokens":524288,"max_output_tokens":512000,"max_tokens":512000,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true,"supports_vision":true,"supports_prompt_caching":true},"openrouter/x-ai/grok-imagine-video":{"provider":"openrouter","base_model":"grok-imagine-video","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/recraft/recraft-v4.1-pro-vector":{"provider":"openrouter","base_model":"recraft-v4.1-pro-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4.1-vector":{"provider":"openrouter","base_model":"recraft-v4.1-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4.1-utility-pro":{"provider":"openrouter","base_model":"recraft-v4.1-utility-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4.1-utility":{"provider":"openrouter","base_model":"recraft-v4.1-utility","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4.1-pro":{"provider":"openrouter","base_model":"recraft-v4.1-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4.1":{"provider":"openrouter","base_model":"recraft-v4.1","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-pro-vector":{"provider":"openrouter","base_model":"recraft-v4-pro-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-vector":{"provider":"openrouter","base_model":"recraft-v4-vector","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4-pro":{"provider":"openrouter","base_model":"recraft-v4-pro","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v4":{"provider":"openrouter","base_model":"recraft-v4","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/recraft/recraft-v3":{"provider":"openrouter","base_model":"recraft-v3","mode":"image_generation","max_input_tokens":65536,"max_output_tokens":58982,"max_tokens":58982,"supports_vision":true},"openrouter/kwaivgi/kling-v3.0-pro":{"provider":"openrouter","base_model":"kling-v3.0-pro","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/kwaivgi/kling-v3.0-std":{"provider":"openrouter","base_model":"kling-v3.0-std","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/google/veo-3.1-fast":{"provider":"openrouter","base_model":"veo-3.1-fast","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/google/veo-3.1-lite":{"provider":"openrouter","base_model":"veo-3.1-lite","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/kwaivgi/kling-video-o1":{"provider":"openrouter","base_model":"kling-video-o1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/minimax/hailuo-2.3":{"provider":"openrouter","base_model":"hailuo-2.3","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/alibaba/wan-2.7":{"provider":"openrouter","base_model":"wan-2.7","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/bytedance/seedance-2.0":{"provider":"openrouter","base_model":"seedance-2.0","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/bytedance/seedance-2.0-fast":{"provider":"openrouter","base_model":"seedance-2.0-fast","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true,"supports_audio_input":true},"openrouter/cohere/rerank-4-pro":{"provider":"openrouter","base_model":"rerank-4-pro","mode":"chat","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491},"openrouter/cohere/rerank-4-fast":{"provider":"openrouter","base_model":"rerank-4-fast","mode":"chat","max_input_tokens":32768,"max_output_tokens":29491,"max_tokens":29491},"openrouter/cohere/rerank-v3.5":{"provider":"openrouter","base_model":"rerank-v3.5","mode":"chat","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686},"openrouter/google/lyria-3-pro-preview":{"provider":"openrouter","base_model":"lyria-3-pro-preview","mode":"chat","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_response_schema":true,"supports_vision":true,"supports_audio_output":true},"openrouter/google/lyria-3-clip-preview":{"provider":"openrouter","base_model":"lyria-3-clip-preview","mode":"chat","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"supports_response_schema":true,"supports_vision":true,"supports_audio_output":true},"openrouter/alibaba/wan-2.6":{"provider":"openrouter","base_model":"wan-2.6","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_vision":true},"openrouter/google/veo-3.1":{"provider":"openrouter","base_model":"veo-3.1","mode":"chat","max_input_tokens":0,"max_output_tokens":0,"max_tokens":0,"supports_response_schema":true,"supports_vision":true},"openrouter/sourceful/riverflow-v2-pro":{"provider":"openrouter","base_model":"riverflow-v2-pro","mode":"image_generation","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_vision":true},"openrouter/sourceful/riverflow-v2-fast":{"provider":"openrouter","base_model":"riverflow-v2-fast","mode":"image_generation","max_input_tokens":8192,"max_output_tokens":7372,"max_tokens":7372,"supports_vision":true},"openrouter/black-forest-labs/flux.2-klein-4b":{"provider":"openrouter","base_model":"flux.2-klein-4b","mode":"image_generation","max_input_tokens":40960,"max_output_tokens":36864,"max_tokens":36864,"supports_vision":true},"openrouter/bytedance-seed/seedream-4.5":{"provider":"openrouter","base_model":"seedream-4.5","mode":"image_generation","max_input_tokens":4096,"max_output_tokens":3686,"max_tokens":3686,"supports_vision":true},"openrouter/black-forest-labs/flux.2-max":{"provider":"openrouter","base_model":"flux.2-max","mode":"image_generation","max_input_tokens":46864,"max_output_tokens":42177,"max_tokens":42177,"supports_vision":true},"openrouter/black-forest-labs/flux.2-flex":{"provider":"openrouter","base_model":"flux.2-flex","mode":"image_generation","max_input_tokens":67344,"max_output_tokens":60609,"max_tokens":60609,"supports_vision":true},"openrouter/black-forest-labs/flux.2-pro":{"provider":"openrouter","base_model":"flux.2-pro","mode":"image_generation","max_input_tokens":46864,"max_output_tokens":42177,"max_tokens":42177,"supports_vision":true},"openrouter/google/gemini-embedding-001:batch":{"provider":"openrouter","base_model":"gemini-embedding-001:batch","mode":"embedding","max_input_tokens":20000,"max_output_tokens":18000,"max_tokens":18000},"openrouter/openai/gpt-oss-120b:batch":{"provider":"openrouter","base_model":"gpt-oss-120b:batch","mode":"chat","max_input_tokens":131072,"max_output_tokens":117964,"max_tokens":117964,"supports_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supports_response_schema":true},"openrouter/morph/morph-v3-large:batch":{"provider":"openrouter","base_model":"morph-v3-large:batch","mode":"chat","max_input_tokens":262144,"max_output_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"openrouter/morph/morph-v3-fast:batch":{"provider":"openrouter","base_model":"morph-v3-fast:batch","mode":"chat","max_input_tokens":81920,"max_output_tokens":38000,"max_tokens":38000},"fireworks_ai/accounts/fireworks/routers/deepseek-v4p1-flash-1017-fast":{"provider":"fireworks_ai","base_model":"deepseek-v4p1-flash-1017-fast","mode":"chat","source":"https://fireworks.ai/pricing","supports_prompt_caching":true,"supports_vision":true,"supports_function_calling":true,"supports_tool_choice":true,"max_input_tokens":1048576,"max_tokens":1048576},"opencode-zen/grok-4.7":{"provider":"opencode-zen","base_model":"grok-4.7","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/claude-haiku-5-5":{"provider":"opencode-zen","base_model":"claude-haiku-5-5","mode":"chat","source":"https://opencode.ai/docs/zen/","supports_native_effort":true},"opencode-zen/gemini-4-argon":{"provider":"opencode-zen","base_model":"gemini-4-argon","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/kimi-k3.1-pro":{"provider":"opencode-zen","base_model":"kimi-k3.1-pro","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/mimo-v2.6-flash":{"provider":"opencode-zen","base_model":"mimo-v2.6-flash","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/mimo-v2.6-pro":{"provider":"opencode-zen","base_model":"mimo-v2.6-pro","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/minimax-m3.1-flash":{"provider":"opencode-zen","base_model":"minimax-m3.1-flash","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/qwen4-max":{"provider":"opencode-zen","base_model":"qwen4-max","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/hy4":{"provider":"opencode-zen","base_model":"hy4","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/step-5-preview":{"provider":"opencode-zen","base_model":"step-5-preview","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/beam-free":{"provider":"opencode-zen","base_model":"beam-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/iceman-contributor-free":{"provider":"opencode-zen","base_model":"iceman-contributor-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/longcat-2.5-free":{"provider":"opencode-zen","base_model":"longcat-2.5-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/hy4-flash-free":{"provider":"opencode-zen","base_model":"hy4-flash-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/claude-fable-5-2":{"provider":"opencode-zen","base_model":"claude-fable-5-2","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/claude-opus-5-5":{"provider":"opencode-zen","base_model":"claude-opus-5-5","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/claude-sonnet-5-5":{"provider":"opencode-zen","base_model":"claude-sonnet-5-5","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/gemini-4-carbon":{"provider":"opencode-zen","base_model":"gemini-4-carbon","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/gemini-4-hydrogen":{"provider":"opencode-zen","base_model":"gemini-4-hydrogen","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/muse-ember-1.0":{"provider":"opencode-zen","base_model":"muse-ember-1.0","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/muse-wonder-1.0":{"provider":"opencode-zen","base_model":"muse-wonder-1.0","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/mistral-large-4":{"provider":"opencode-zen","base_model":"mistral-large-4","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/gpt-6-luna-decision":{"provider":"opencode-zen","base_model":"gpt-6-luna-decision","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/gemini-decision-1":{"provider":"opencode-zen","base_model":"gemini-decision-1","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/jev-1.13":{"provider":"opencode-zen","base_model":"jev-1.13","mode":"decisions","source":"https://opencode.ai/docs/zen/","supports_decisions":true},"opencode-zen/jev-1.13-free":{"provider":"opencode-zen","base_model":"jev-1.13-free","mode":"decisions","source":"https://opencode.ai/docs/zen/","supports_decisions":true},"opencode-zen/iceman-free":{"provider":"opencode-zen","base_model":"iceman-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/inkling-2-small-free":{"provider":"opencode-zen","base_model":"inkling-2-small-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/glm-5.3-flash-free":{"provider":"opencode-zen","base_model":"glm-5.3-flash-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/deepseek-v4.1-flash-free":{"provider":"opencode-zen","base_model":"deepseek-v4.1-flash-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/minimax-m3.1-flash-free":{"provider":"opencode-zen","base_model":"minimax-m3.1-flash-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/flint-40b-free":{"provider":"opencode-zen","base_model":"flint-40b-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/mimo-v2.6-flash-free":{"provider":"opencode-zen","base_model":"mimo-v2.6-flash-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"opencode-zen/nemotron-3.5-super-vl-free":{"provider":"opencode-zen","base_model":"nemotron-3.5-super-vl-free","mode":"chat","source":"https://opencode.ai/docs/zen/"},"huggingface/akashml/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/dekallm/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/io-net/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/ionstream/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/mancer/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modelrun/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/near-ai/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/reka/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/tenstorrent/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nscale/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/ovhcloud/Qwen/Qwen3.8-27B":{"provider":"huggingface","base_model":"Qwen3.8-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/deepinfra/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/thinkingmachines/Inkling-2-Small":{"provider":"huggingface","base_model":"Inkling-2-Small","mode":"chat","source":"https://huggingface.co/thinkingmachines/Inkling-2-Small","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/akashml/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/decart/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/deepinfra/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/fireworks-ai/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/moonshotai/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/morph/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primalabs/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/relace/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/river/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/sail-research/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/moonshotai/Kimi-K3.1-Pro":{"provider":"huggingface","base_model":"Kimi-K3.1-Pro","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3.1-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/baseten/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/cloudflare/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cohere/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/dekallm/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/fireworks-ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1040000,"max_tokens":1040000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/ionstream/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/makora/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/mistral/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modelrun/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/morph/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/near-ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/novita/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/open-inference/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/particle-ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primalabs/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/relace/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/river/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/sail-research/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/stepfun/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash-1017","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/liquid/LiquidAI/LFM3-1.7B":{"provider":"huggingface","base_model":"LFM3-1.7B","mode":"chat","source":"https://huggingface.co/LiquidAI/LFM3-1.7B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/meta-llama/Llama-3.1-8B-Instruct":{"provider":"huggingface","base_model":"Llama-3.1-8B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":32000,"max_tokens":32000},"huggingface/coreweave/meta-llama/Llama-3.1-8B-Instruct":{"provider":"huggingface","base_model":"Llama-3.1-8B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/groq/meta-llama/Llama-3.1-8B-Instruct":{"provider":"huggingface","base_model":"Llama-3.1-8B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/cloudflare/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cohere/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/fireworks-ai/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1040000,"max_tokens":1040000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/makora/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/mancer/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/open-inference/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/reka/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/relace/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/river/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/together/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/deepseek-ai/DeepSeek-V4.1-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4.1-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/akashml/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/amazon-bedrock/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cohere/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/coreweave/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/crusoe/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/decart/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/doubleword/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/featherless-ai/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/fireworks-ai/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/inceptron/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/inference-net/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/makora/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mistral/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/modal/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/modelrun/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/morph/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/relace/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/sail-research/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/sambanova/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/tenstorrent/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1044480,"max_tokens":1044480,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/wafer/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nscale/zai-org/GLM-5.3":{"provider":"huggingface","base_model":"GLM-5.3","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/coreweave/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modelrun/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nextbit/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/google/gemma-4-31B-it":{"provider":"huggingface","base_model":"gemma-4-31B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-31B-it","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cohere/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/dekallm/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/fireworks-ai/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inceptron/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/makora/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/mistral/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/morph/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/near-ai/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nextbit/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/open-inference/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/particle-ai/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/reka/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/relace/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/river/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/sail-research/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/tenstorrent/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":262000,"max_tokens":262000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/zai-org/zai-org/GLM-5.3-Flash":{"provider":"huggingface","base_model":"GLM-5.3-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.3-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/Qwen/Qwen3.5-9B":{"provider":"huggingface","base_model":"Qwen3.5-9B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-9B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inferact-vllm/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/makora/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modelrun/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/sambanova/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/novita/MiniMaxAI/MiniMax-M3.1-Flash":{"provider":"huggingface","base_model":"MiniMax-M3.1-Flash","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/amazon-bedrock/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/digitalocean/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inference-net/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/io-net/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048320,"max_tokens":1048320,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/makora/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/moonshotai/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/relace/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/river/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_response_schema":true,"supports_vision":true},"huggingface/wafer/moonshotai/Kimi-K3":{"provider":"huggingface","base_model":"Kimi-K3","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/akashml/Qwen/Qwen3.6-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.6-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/Qwen/Qwen3.6-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.6-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3.6-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.6-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/Qwen/Qwen3.6-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.6-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3.6-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.6-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/modelrun/meta-models/Muse-Glimmer-30B":{"provider":"huggingface","base_model":"Muse-Glimmer-30B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/meta-llama/Llama-3.2-3B-Instruct":{"provider":"huggingface","base_model":"Llama-3.2-3B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct","max_input_tokens":80000,"max_tokens":80000},"huggingface/parasail/meta-llama/Llama-3.2-3B-Instruct":{"provider":"huggingface","base_model":"Llama-3.2-3B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/akashml/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/amazon-bedrock/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/coreweave/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/crusoe/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/dekallm/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":128000,"max_tokens":128000},"huggingface/google-vertex/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/mancer/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/near-ai/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/primeintellect/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/sail-research/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/sambanova/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/venice/openai/gpt-oss-120b":{"provider":"huggingface","base_model":"gpt-oss-120b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-120b","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/akashml/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/coreweave/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/dekallm/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/google-vertex/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/io-net/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262142,"max_tokens":262142,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/makora/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nextbit/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/google/gemma-4-26B-A4B-it":{"provider":"huggingface","base_model":"gemma-4-26B-A4B-it","mode":"chat","source":"https://huggingface.co/google/gemma-4-26B-A4B-it","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/akashml/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/amazon-bedrock/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/coreweave/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/dekallm/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/primeintellect/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/openai/gpt-oss-20b":{"provider":"huggingface","base_model":"gpt-oss-20b","mode":"chat","source":"https://huggingface.co/openai/gpt-oss-20b","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/parasail/google/gemma-3-4b-it":{"provider":"huggingface","base_model":"gemma-3-4b-it","mode":"chat","source":"https://huggingface.co/google/gemma-3-4b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true},"huggingface/mistral/mistralai/Mistral-Large-4-1T-A52B":{"provider":"huggingface","base_model":"Mistral-Large-4-1T-A52B","mode":"chat","source":"https://huggingface.co/mistralai/Mistral-Large-4-1T-A52B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/Qwen/Qwen3.8-2.4T-A95B":{"provider":"huggingface","base_model":"Qwen3.8-2.4T-A95B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/Qwen/Qwen3.8-2.4T-A95B":{"provider":"huggingface","base_model":"Qwen3.8-2.4T-A95B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/azure/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/fireworks-ai/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mancer/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/relace/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/wafer/deepseek-ai/DeepSeek-V4-Flash":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/baseten/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nscale/Qwen/Qwen3.8-Flash-Next-FP8":{"provider":"huggingface","base_model":"Qwen3.8-Flash-Next-FP8","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next-FP8","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/amazon-bedrock/Qwen/Qwen3-Coder-30B-A3B-Instruct":{"provider":"huggingface","base_model":"Qwen3-Coder-30B-A3B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/deepinfra/tencent/Hy4-preview":{"provider":"huggingface","base_model":"Hy4-preview","mode":"chat","source":"https://huggingface.co/tencent/Hy4-preview","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/tencent/Hy4-preview":{"provider":"huggingface","base_model":"Hy4-preview","mode":"chat","source":"https://huggingface.co/tencent/Hy4-preview","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/tencent/tencent/Hy4-preview":{"provider":"huggingface","base_model":"Hy4-preview","mode":"chat","source":"https://huggingface.co/tencent/Hy4-preview","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/tencent/Hy4-preview":{"provider":"huggingface","base_model":"Hy4-preview","mode":"chat","source":"https://huggingface.co/tencent/Hy4-preview","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cohere/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/coreweave/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/inceptron/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mancer/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/primeintellect/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/reka/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/relace/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/sail-research/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/wafer/deepseek-ai/DeepSeek-V4-Flash-0731":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-0731","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/liquid/LiquidAI/LFM2.5-24B-A2B":{"provider":"huggingface","base_model":"LFM2.5-24B-A2B","mode":"chat","source":"https://huggingface.co/LiquidAI/LFM2.5-24B-A2B","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/baseten/meta-models/Muse-Spark-Open-816B-A42B":{"provider":"huggingface","base_model":"Muse-Spark-Open-816B-A42B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/fireworks-ai/meta-models/Muse-Spark-Open-816B-A42B":{"provider":"huggingface","base_model":"Muse-Spark-Open-816B-A42B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/modal/meta-models/Muse-Spark-Open-816B-A42B":{"provider":"huggingface","base_model":"Muse-Spark-Open-816B-A42B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/meta-models/Muse-Spark-Open-816B-A42B":{"provider":"huggingface","base_model":"Muse-Spark-Open-816B-A42B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/meta-models/Muse-Spark-Open-816B-A42B":{"provider":"huggingface","base_model":"Muse-Spark-Open-816B-A42B","mode":"chat","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/akashml/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":24000,"max_tokens":24000},"huggingface/coreweave/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/deepinfra/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/google-vertex/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/groq/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/parasail/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/sambanova-turbo/meta-llama/Llama-3.3-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.3-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/azure/deepseek-ai/DeepSeek-V4-Pro":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V4-Pro":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/relace/deepseek-ai/DeepSeek-V4-Pro":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4-Pro":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/deepseek-ai/DeepSeek-V4-Pro":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/deepinfra/XiaomiMiMo/MiMo-V2.6-Pro-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Pro-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/primalabs/XiaomiMiMo/MiMo-V2.6-Pro-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Pro-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/novita/XiaomiMiMo/MiMo-V2.6-Pro-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Pro-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/parasail/Qwen/Qwen3-VL-8B-Instruct":{"provider":"huggingface","base_model":"Qwen3-VL-8B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-VL-8B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/novita/inclusionAI/Ling-3.1-flash":{"provider":"huggingface","base_model":"Ling-3.1-flash","mode":"chat","source":"https://huggingface.co/inclusionAI/Ling-3.1-flash","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/cloudflare/meta-llama/Llama-3.2-1B-Instruct":{"provider":"huggingface","base_model":"Llama-3.2-1B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct","max_input_tokens":60000,"max_tokens":60000},"huggingface/cohere/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/deepinfra/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/makora/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/tenstorrent/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/novita/XiaomiMiMo/MiMo-V2.6-Flash-RL":{"provider":"huggingface","base_model":"MiMo-V2.6-Flash-RL","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/baseten/reflection/Beam-1.0-501B-A23B":{"provider":"huggingface","base_model":"Beam-1.0-501B-A23B","mode":"chat","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","max_input_tokens":1048576,"max_tokens":1048576,"supports_response_schema":true},"huggingface/coreweave/reflection/Beam-1.0-501B-A23B":{"provider":"huggingface","base_model":"Beam-1.0-501B-A23B","mode":"chat","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/reflection/Beam-1.0-501B-A23B":{"provider":"huggingface","base_model":"Beam-1.0-501B-A23B","mode":"chat","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/reflection/Beam-1.0-501B-A23B":{"provider":"huggingface","base_model":"Beam-1.0-501B-A23B","mode":"chat","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/together/reflection/Beam-1.0-501B-A23B":{"provider":"huggingface","base_model":"Beam-1.0-501B-A23B","mode":"chat","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/baseten/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/coreweave/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/deepinfra/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/io-net/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_vision":true},"huggingface/nebius/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/together/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/relace/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/sail-research/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/wafer/deepseek-ai/DeepSeek-V4-Pro-0813":{"provider":"huggingface","base_model":"DeepSeek-V4-Pro-0813","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/coreweave/MiniMaxAI/MiniMax-M3":{"provider":"huggingface","base_model":"MiniMax-M3","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/MiniMaxAI/MiniMax-M3":{"provider":"huggingface","base_model":"MiniMax-M3","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/venice/MiniMaxAI/MiniMax-M3":{"provider":"huggingface","base_model":"MiniMax-M3","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M3","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/nextbit/Qwen/Qwen3-14B":{"provider":"huggingface","base_model":"Qwen3-14B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-14B","max_input_tokens":40960,"max_tokens":40960,"supports_response_schema":true},"huggingface/siliconflow/Qwen/Qwen3.5-122B-A10B":{"provider":"huggingface","base_model":"Qwen3.5-122B-A10B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-122B-A10B","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/deepseek-ai/DeepSeek-V3.2":{"provider":"huggingface","base_model":"DeepSeek-V3.2","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2","max_input_tokens":163840,"max_tokens":163840},"huggingface/siliconflow/deepseek-ai/DeepSeek-V3.2":{"provider":"huggingface","base_model":"DeepSeek-V3.2","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/zai-org/GLM-4.7-Flash":{"provider":"huggingface","base_model":"GLM-4.7-Flash","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.7-Flash","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/Qwen/Qwen2.5-Coder-32B-Instruct":{"provider":"huggingface","base_model":"Qwen2.5-Coder-32B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct","max_input_tokens":32768,"max_tokens":32768},"huggingface/coreweave/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/deepinfra/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/fireworks-ai/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/modelrun/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/cloudflare/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/coreweave/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/inceptron/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mistral/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/morph/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/relace/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/wafer/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-5.2":{"provider":"huggingface","base_model":"GLM-5.2","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.2","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"provider":"huggingface","base_model":"DeepSeek-V4-Flash-Vision-Exp","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","max_input_tokens":1048576,"max_tokens":1048576,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/venice/dphn/Dolphin-Mistral-24B-Venice-Edition":{"provider":"huggingface","base_model":"Dolphin-Mistral-24B-Venice-Edition","mode":"chat","source":"https://huggingface.co/dphn/Dolphin-Mistral-24B-Venice-Edition","max_input_tokens":128000,"max_tokens":128000},"huggingface/nextbit/google/gemma-3-12b-it":{"provider":"huggingface","base_model":"gemma-3-12b-it","mode":"chat","source":"https://huggingface.co/google/gemma-3-12b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3-Coder-Next":{"provider":"huggingface","base_model":"Qwen3-Coder-Next","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-Coder-Next","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/google-vertex/meta-llama/Llama-4-Maverick-17B-128E-Instruct":{"provider":"huggingface","base_model":"Llama-4-Maverick-17B-128E-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/meta-llama/Llama-4-Maverick-17B-128E-Instruct":{"provider":"huggingface","base_model":"Llama-4-Maverick-17B-128E-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct","max_input_tokens":524288,"max_tokens":524288,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/reka/Qwen/Qwen3.6-27B":{"provider":"huggingface","base_model":"Qwen3.6-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/Qwen/Qwen3.6-27B":{"provider":"huggingface","base_model":"Qwen3.6-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3.6-27B":{"provider":"huggingface","base_model":"Qwen3.6-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-27B","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/ovhcloud/Qwen/Qwen3.6-27B":{"provider":"huggingface","base_model":"Qwen3.6-27B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.6-27B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/crusoe/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/digitalocean/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inceptron/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/moonshotai/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/moonshotai/Kimi-K2.6":{"provider":"huggingface","base_model":"Kimi-K2.6","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.6","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nextbit/google/gemma-2-27b-it":{"provider":"huggingface","base_model":"gemma-2-27b-it","mode":"chat","source":"https://huggingface.co/google/gemma-2-27b-it","max_input_tokens":8192,"max_tokens":8192,"supports_response_schema":true},"huggingface/baseten/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","max_input_tokens":202800,"max_tokens":202800,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/google-vertex/meta-llama/Llama-4-Scout-17B-16E-Instruct":{"provider":"huggingface","base_model":"Llama-4-Scout-17B-16E-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct","max_input_tokens":1310720,"max_tokens":1310720,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/novita/tencent/Hy3":{"provider":"huggingface","base_model":"Hy3","mode":"chat","source":"https://huggingface.co/tencent/Hy3","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/tencent/tencent/Hy3":{"provider":"huggingface","base_model":"Hy3","mode":"chat","source":"https://huggingface.co/tencent/Hy3","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/ByteDance-Seed/UI-TARS-1.5-7B":{"provider":"huggingface","base_model":"UI-TARS-1.5-7B","mode":"chat","source":"https://huggingface.co/ByteDance-Seed/UI-TARS-1.5-7B","max_input_tokens":128000,"max_tokens":128000,"supports_response_schema":true,"supports_vision":true},"huggingface/primeintellect/moondream/moondream3.1-9B-A2B":{"provider":"huggingface","base_model":"moondream3.1-9B-A2B","mode":"chat","source":"https://huggingface.co/moondream/moondream3.1-9B-A2B","max_input_tokens":32000,"max_tokens":32000,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3.5-397B-A17B":{"provider":"huggingface","base_model":"Qwen3.5-397B-A17B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3.5-397B-A17B":{"provider":"huggingface","base_model":"Qwen3.5-397B-A17B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/nebius/google/gemma-3-27b-it":{"provider":"huggingface","base_model":"gemma-3-27b-it","mode":"chat","source":"https://huggingface.co/google/gemma-3-27b-it","max_input_tokens":110000,"max_tokens":110000,"supports_response_schema":true,"supports_vision":true},"huggingface/nextbit/google/gemma-3-27b-it":{"provider":"huggingface","base_model":"gemma-3-27b-it","mode":"chat","source":"https://huggingface.co/google/gemma-3-27b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/google/gemma-3-27b-it":{"provider":"huggingface","base_model":"gemma-3-27b-it","mode":"chat","source":"https://huggingface.co/google/gemma-3-27b-it","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true,"supports_vision":true},"huggingface/cloudflare/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/inceptron/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/moonshotai/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/venice/moonshotai/Kimi-K2.7-Code":{"provider":"huggingface","base_model":"Kimi-K2.7-Code","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3.5-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.5-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/Qwen/Qwen3.5-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.5-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B","max_input_tokens":262144,"max_tokens":262144,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3.5-35B-A3B":{"provider":"huggingface","base_model":"Qwen3.5-35B-A3B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/amazon-bedrock/meta-llama/Llama-3.1-70B-Instruct":{"provider":"huggingface","base_model":"Llama-3.1-70B-Instruct","mode":"chat","source":"https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/digitalocean/MiniMaxAI/MiniMax-M2.5":{"provider":"huggingface","base_model":"MiniMax-M2.5","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.5","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/MiniMaxAI/MiniMax-M2.5":{"provider":"huggingface","base_model":"MiniMax-M2.5","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.5","max_input_tokens":198000,"max_tokens":198000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/novita/moonshotai/Kimi-K2-Thinking":{"provider":"huggingface","base_model":"Kimi-K2-Thinking","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2-Thinking","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/zai-org/GLM-4.5-Air":{"provider":"huggingface","base_model":"GLM-4.5-Air","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.5-Air","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/zai-org/zai-org/GLM-4.5-Air":{"provider":"huggingface","base_model":"GLM-4.5-Air","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.5-Air","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/primeintellect/moondream/moondream3-preview":{"provider":"huggingface","base_model":"moondream3-preview","mode":"chat","source":"https://huggingface.co/moondream/moondream3-preview","max_input_tokens":32000,"max_tokens":32000,"supports_vision":true},"huggingface/amazon-bedrock/moonshotai/Kimi-K2.5":{"provider":"huggingface","base_model":"Kimi-K2.5","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.5","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/siliconflow/moonshotai/Kimi-K2.5":{"provider":"huggingface","base_model":"Kimi-K2.5","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.5","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/moonshotai/Kimi-K2.5":{"provider":"huggingface","base_model":"Kimi-K2.5","mode":"chat","source":"https://huggingface.co/moonshotai/Kimi-K2.5","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/Qwen/Qwen3-32B":{"provider":"huggingface","base_model":"Qwen3-32B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-32B","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/tenstorrent/Qwen/Qwen3-32B":{"provider":"huggingface","base_model":"Qwen3-32B","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-32B","max_input_tokens":20000,"max_tokens":20000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/Qwen/Qwen3-VL-30B-A3B-Instruct":{"provider":"huggingface","base_model":"Qwen3-VL-30B-A3B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-VL-30B-A3B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/coreweave/deepseek-ai/DeepSeek-V3.1":{"provider":"huggingface","base_model":"DeepSeek-V3.1","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1","max_input_tokens":161000,"max_tokens":161000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V3.1":{"provider":"huggingface","base_model":"DeepSeek-V3.1","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/Qwen/Qwen3-Next-80B-A3B-Instruct":{"provider":"huggingface","base_model":"Qwen3-Next-80B-A3B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mancer/Gryphe/MythoMax-L2-13b":{"provider":"huggingface","base_model":"MythoMax-L2-13b","mode":"chat","source":"https://huggingface.co/Gryphe/MythoMax-L2-13b","max_input_tokens":8192,"max_tokens":8192,"supports_response_schema":true},"huggingface/parasail/Gryphe/MythoMax-L2-13b":{"provider":"huggingface","base_model":"MythoMax-L2-13b","mode":"chat","source":"https://huggingface.co/Gryphe/MythoMax-L2-13b","max_input_tokens":4096,"max_tokens":4096,"supports_response_schema":true},"huggingface/novita/inclusionAI/Ling-3.0-flash-Fin":{"provider":"huggingface","base_model":"Ling-3.0-flash-Fin","mode":"chat","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"provider":"huggingface","base_model":"Qwen3-Coder-480B-A35B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"provider":"huggingface","base_model":"Qwen3-Coder-480B-A35B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/dekallm/Qwen/Qwen3-30B-A3B-Instruct-2507":{"provider":"huggingface","base_model":"Qwen3-30B-A3B-Instruct-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/Qwen/Qwen3-30B-A3B-Instruct-2507":{"provider":"huggingface","base_model":"Qwen3-30B-A3B-Instruct-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/siliconflow/Qwen/Qwen3-30B-A3B-Instruct-2507":{"provider":"huggingface","base_model":"Qwen3-30B-A3B-Instruct-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/groq/MiniMaxAI/MiniMax-M2.7":{"provider":"huggingface","base_model":"MiniMax-M2.7","mode":"chat","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.7","max_input_tokens":196608,"max_tokens":196608,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/publicai/aisingapore/Qwen-SEA-LION-v4.5-27B-IT":{"provider":"huggingface","base_model":"Qwen-SEA-LION-v4.5-27B-IT","mode":"chat","source":"https://huggingface.co/aisingapore/Qwen-SEA-LION-v4.5-27B-IT","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-R1-0528":{"provider":"huggingface","base_model":"DeepSeek-R1-0528","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/novita/inclusionAI/Ling-3.0-flash-VL":{"provider":"huggingface","base_model":"Ling-3.0-flash-VL","mode":"chat","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-VL","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/mancer/Undi95/ReMM-SLERP-L2-13B":{"provider":"huggingface","base_model":"ReMM-SLERP-L2-13B","mode":"chat","source":"https://huggingface.co/Undi95/ReMM-SLERP-L2-13B","max_input_tokens":6144,"max_tokens":6144,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-4.5V":{"provider":"huggingface","base_model":"GLM-4.5V","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.5V","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_vision":true},"huggingface/deepinfra/Sao10K/L3-8B-Lunaris-v1":{"provider":"huggingface","base_model":"L3-8B-Lunaris-v1","mode":"chat","source":"https://huggingface.co/Sao10K/L3-8B-Lunaris-v1","max_input_tokens":8192,"max_tokens":8192},"huggingface/parasail/Sao10K/L3-8B-Lunaris-v1":{"provider":"huggingface","base_model":"L3-8B-Lunaris-v1","mode":"chat","source":"https://huggingface.co/Sao10K/L3-8B-Lunaris-v1","max_input_tokens":8192,"max_tokens":8192,"supports_response_schema":true},"huggingface/reka/RekaAI/reka-edge-2603":{"provider":"huggingface","base_model":"reka-edge-2603","mode":"chat","source":"https://huggingface.co/RekaAI/reka-edge-2603","max_input_tokens":16384,"max_tokens":16384,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/mancer/anthracite-org/magnum-v4-72b":{"provider":"huggingface","base_model":"magnum-v4-72b","mode":"chat","source":"https://huggingface.co/anthracite-org/magnum-v4-72b","max_input_tokens":32768,"max_tokens":32768,"supports_response_schema":true},"huggingface/venice/zai-org/GLM-4.6":{"provider":"huggingface","base_model":"GLM-4.6","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.6","max_input_tokens":198000,"max_tokens":198000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-4.6":{"provider":"huggingface","base_model":"GLM-4.6","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.6","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/cohere/CohereLabs/command-a-plus-05-2026-bf16":{"provider":"huggingface","base_model":"command-a-plus-05-2026-bf16","mode":"chat","source":"https://huggingface.co/CohereLabs/command-a-plus-05-2026-bf16","max_input_tokens":192000,"max_tokens":192000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V3-0324":{"provider":"huggingface","base_model":"DeepSeek-V3-0324","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3-0324","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/mancer/zai-org/GLM-4.7":{"provider":"huggingface","base_model":"GLM-4.7","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":131072,"max_tokens":131072},"huggingface/venice/zai-org/GLM-4.7":{"provider":"huggingface","base_model":"GLM-4.7","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":198000,"max_tokens":198000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-4.7":{"provider":"huggingface","base_model":"GLM-4.7","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.7","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/novita/stepfun-ai/Step-3.7-Flash":{"provider":"huggingface","base_model":"Step-3.7-Flash","mode":"chat","source":"https://huggingface.co/stepfun-ai/Step-3.7-Flash","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/stepfun/stepfun-ai/Step-3.7-Flash":{"provider":"huggingface","base_model":"Step-3.7-Flash","mode":"chat","source":"https://huggingface.co/stepfun-ai/Step-3.7-Flash","max_input_tokens":256000,"max_tokens":256000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen3-235B-A22B-Instruct-2507":{"provider":"huggingface","base_model":"Qwen3-235B-A22B-Instruct-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/Qwen/Qwen3-235B-A22B-Instruct-2507":{"provider":"huggingface","base_model":"Qwen3-235B-A22B-Instruct-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/amazon-bedrock/zai-org/GLM-5":{"provider":"huggingface","base_model":"GLM-5","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/zai-org/GLM-5":{"provider":"huggingface","base_model":"GLM-5","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/zai-org/GLM-5":{"provider":"huggingface","base_model":"GLM-5","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":198000,"max_tokens":198000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-5":{"provider":"huggingface","base_model":"GLM-5","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/Qwen/Qwen3-235B-A22B-Thinking-2507":{"provider":"huggingface","base_model":"Qwen3-235B-A22B-Thinking-2507","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/zai-org/zai-org/GLM-4.5":{"provider":"huggingface","base_model":"GLM-4.5","mode":"chat","source":"https://huggingface.co/zai-org/GLM-4.5","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V3.1-Terminus":{"provider":"huggingface","base_model":"DeepSeek-V3.1-Terminus","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1-Terminus","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/nebius/NousResearch/Hermes-4-405B":{"provider":"huggingface","base_model":"Hermes-4-405B","mode":"chat","source":"https://huggingface.co/NousResearch/Hermes-4-405B","max_input_tokens":131072,"max_tokens":131072,"supports_response_schema":true},"huggingface/siliconflow/deepseek-ai/DeepSeek-V3.2-Exp":{"provider":"huggingface","base_model":"DeepSeek-V3.2-Exp","mode":"chat","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp","max_input_tokens":163840,"max_tokens":163840,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/novita/zai-org/GLM-5.1":{"provider":"huggingface","base_model":"GLM-5.1","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.1","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/siliconflow/zai-org/GLM-5.1":{"provider":"huggingface","base_model":"GLM-5.1","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.1","max_input_tokens":204800,"max_tokens":204800,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/venice/zai-org/GLM-5.1":{"provider":"huggingface","base_model":"GLM-5.1","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.1","max_input_tokens":200000,"max_tokens":200000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/zai-org/zai-org/GLM-5.1":{"provider":"huggingface","base_model":"GLM-5.1","mode":"chat","source":"https://huggingface.co/zai-org/GLM-5.1","max_input_tokens":202752,"max_tokens":202752,"supports_function_calling":true,"supports_tool_choice":true},"huggingface/nebius/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3-Nano-30B-A3B-BF16","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/digitalocean/XiaomiMiMo/MiMo-V2.5-Pro":{"provider":"huggingface","base_model":"MiMo-V2.5-Pro","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/deepinfra/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3-Super-120B-A12B-FP8","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/dekallm/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8":{"provider":"huggingface","base_model":"NVIDIA-Nemotron-3-Super-120B-A12B-FP8","mode":"chat","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/venice/XiaomiMiMo/MiMo-V2.5":{"provider":"huggingface","base_model":"MiMo-V2.5","mode":"chat","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5","max_input_tokens":1000000,"max_tokens":1000000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/parasail/Qwen/Qwen3-VL-235B-A22B-Instruct":{"provider":"huggingface","base_model":"Qwen3-VL-235B-A22B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct","max_input_tokens":131072,"max_tokens":131072,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/venice/Qwen/Qwen3-VL-235B-A22B-Instruct":{"provider":"huggingface","base_model":"Qwen3-VL-235B-A22B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct","max_input_tokens":128000,"max_tokens":128000,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true,"supports_vision":true},"huggingface/parasail/Qwen/Qwen2.5-VL-72B-Instruct":{"provider":"huggingface","base_model":"Qwen2.5-VL-72B-Instruct","mode":"chat","source":"https://huggingface.co/Qwen/Qwen2.5-VL-72B-Instruct","max_input_tokens":128000,"max_tokens":128000,"supports_response_schema":true,"supports_vision":true},"huggingface/publicai/BSC-LT/ALIA-40b-instruct-2606":{"provider":"huggingface","base_model":"ALIA-40b-instruct-2606","mode":"chat","source":"https://huggingface.co/BSC-LT/ALIA-40b-instruct-2606","max_input_tokens":65536,"max_tokens":65536,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"huggingface/publicai/aisingapore/Nemotron-SEA-LION-v4.8-120B-A12B":{"provider":"huggingface","base_model":"Nemotron-SEA-LION-v4.8-120B-A12B","mode":"chat","source":"https://huggingface.co/aisingapore/Nemotron-SEA-LION-v4.8-120B-A12B","max_input_tokens":262144,"max_tokens":262144,"supports_function_calling":true,"supports_tool_choice":true,"supports_response_schema":true},"together_ai/black-forest-labs/FLUX.1-kontext-max":{"provider":"together_ai","base_model":"flux.1-kontext-max","mode":"image_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/black-forest-labs/FLUX.1-kontext-pro":{"provider":"together_ai","base_model":"flux.1-kontext-pro","mode":"image_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/black-forest-labs/FLUX.1.1-pro":{"provider":"together_ai","base_model":"flux.1.1-pro","mode":"image_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/black-forest-labs/FLUX.2-max":{"provider":"together_ai","base_model":"flux.2-max","mode":"image_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/ByteDance/Seedance-1.0-lite":{"provider":"together_ai","base_model":"seedance-1.0-lite","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/ByteDance/Seedance-1.0-pro":{"provider":"together_ai","base_model":"seedance-1.0-pro","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/minimax/video-01-director":{"provider":"together_ai","base_model":"video-01-director","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/MiniMaxAI/MiniMax-H3":{"provider":"together_ai","base_model":"minimax-h3","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/minimax/hailuo-02":{"provider":"together_ai","base_model":"hailuo-02","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/pixverse/pixverse-v5":{"provider":"together_ai","base_model":"pixverse-v5","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/vidu/vidu-q1":{"provider":"together_ai","base_model":"vidu-q1","mode":"video_generation","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/canopylabs/orpheus-3b-0.1-ft":{"provider":"together_ai","base_model":"orpheus-3b-0.1-ft","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/cartesia/sonic":{"provider":"together_ai","base_model":"sonic","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/cartesia/sonic-2":{"provider":"together_ai","base_model":"sonic-2","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/cartesia/sonic-3":{"provider":"together_ai","base_model":"sonic-3","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/hexgrad/Kokoro-82M":{"provider":"together_ai","base_model":"kokoro-82m","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/nvidia/nemotron-3-asr-streaming-0.6b":{"provider":"together_ai","base_model":"nemotron-3-asr-streaming-0.6b","mode":"audio_transcription","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/nvidia/nemotron-3.5-asr-streaming-0.6b":{"provider":"together_ai","base_model":"nemotron-3.5-asr-streaming-0.6b","mode":"audio_transcription","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/nvidia/parakeet-tdt-0.6b-v3":{"provider":"together_ai","base_model":"parakeet-tdt-0.6b-v3","mode":"audio_transcription","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/openai/whisper-large-v3":{"provider":"together_ai","base_model":"whisper-large-v3","mode":"audio_transcription","source":"https://docs.together.ai/docs/serverless/models.md"},"together_ai/Qwen/Qwen3-TTS-12Hz-1.7B":{"provider":"together_ai","base_model":"qwen3-tts-12hz-1.7b","mode":"audio_speech","source":"https://docs.together.ai/docs/serverless/models.md"},"wafer/Kimi-K3.1-Pro":{"provider":"wafer","base_model":"kimi-k3.1-pro","mode":"chat","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"wafer/Kimi-K3.1-Pro-US":{"provider":"wafer","base_model":"kimi-k3.1-pro-us","mode":"chat","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"wafer/MiniMax-M3.1-Flash":{"provider":"wafer","base_model":"minimax-m3.1-flash","mode":"chat","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"wafer/DeepSeek-V4-Flash-0423":{"provider":"wafer","base_model":"deepseek-v4-flash-0423","mode":"chat","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":false,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"wafer/DeepSeek-V4.1-Flash-1017":{"provider":"wafer","base_model":"deepseek-v4.1-flash-1017","mode":"chat","source":"https://pass.wafer.ai/v1/models","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"supports_vision":true,"supports_function_calling":true,"supports_reasoning":true,"supports_response_schema":true},"kimchi/deepseek-v4-flash-0731":{"provider":"kimchi","base_model":"deepseek-v4-flash","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":512000,"max_tokens":512000,"supported_modalities":["text"],"supports_vision":false,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"deepseek-v4-flash-0731"},"kimchi/deepseek-v4.1-flash":{"provider":"kimchi","base_model":"deepseek-v4.1-flash","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":512000,"max_tokens":512000,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"is_deprecated":true,"deprecation_date":"2026-12-02","model":"deepseek-v4.1-flash","comment":"Deprecated 2026-11-02; sunset 2026-12-02. Replacement: kimchi/deepseek-v4.1-flash-1017."},"kimchi/deepseek-v4.1-flash-1017":{"provider":"kimchi","base_model":"deepseek-v4.1-flash","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":512000,"max_tokens":512000,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"deepseek-v4.1-flash-1017"},"kimchi/gemma-4-31b":{"provider":"kimchi","base_model":"gemma-4-31b","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"gemma-4-31b"},"kimchi/glm-5.3":{"provider":"kimchi","base_model":"glm-5.3","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text"],"supports_vision":false,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"glm-5.3"},"kimchi/glm-5.3-flash":{"provider":"kimchi","base_model":"glm-5.3-flash","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"glm-5.3-flash"},"kimchi/kimi-k3":{"provider":"kimchi","base_model":"kimi-k3","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":60000,"max_tokens":60000,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"kimi-k3"},"kimchi/kimi-k3.1-pro":{"provider":"kimchi","base_model":"kimi-k3.1-pro","mode":"chat","source":"https://app.kimchi.dev/api/ai-optimizer/v1beta/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true},"kimchi/minimax-m3":{"provider":"kimchi","base_model":"minimax-m3","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":524288,"max_tokens":524288,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"minimax-m3"},"kimchi/minimax-m3.1-flash":{"provider":"kimchi","base_model":"minimax-m3.1-flash","mode":"chat","source":"https://app.kimchi.dev/api/ai-optimizer/v1beta/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":524288,"max_tokens":524288,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true},"kimchi/muse-spark-open-fp4":{"provider":"kimchi","base_model":"muse-spark-open-fp4","mode":"chat","source":"https://app.kimchi.dev/api/ai-optimizer/v1beta/model-prices","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_modalities":["text","image"],"supports_vision":true,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true},"kimchi/nemotron-3-ultra-fp4":{"provider":"kimchi","base_model":"nemotron-3-ultra","mode":"chat","source":"https://llm.kimchi.dev/v1/model-prices","max_input_tokens":1048576,"max_output_tokens":66000,"max_tokens":66000,"supported_modalities":["text"],"supports_vision":false,"supported_output_modalities":["text"],"supports_function_calling":true,"supports_reasoning":true,"model":"nemotron-3-ultra-fp4","supports_prompt_caching":false},"databricks/system.ai.gpt-6-1-terra":{"provider":"databricks","base_model":"databricks-gpt-6-1-terra","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gemini-4-argon":{"provider":"databricks","base_model":"databricks-gemini-4-argon","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":1048576,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gpt-6-1-sol":{"provider":"databricks","base_model":"databricks-gpt-6-1-sol","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-6-sol":{"provider":"databricks","base_model":"databricks-gpt-6-sol","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-6-luna":{"provider":"databricks","base_model":"databricks-gpt-6-luna","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-6-astra":{"provider":"databricks","base_model":"databricks-gpt-6-astra","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-6-sol":{"provider":"databricks","base_model":"databricks-gpt-5-6-sol","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-6-terra":{"provider":"databricks","base_model":"databricks-gpt-5-6-terra","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-6-luna":{"provider":"databricks","base_model":"databricks-gpt-5-6-luna","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":922000,"max_tokens":922000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-5-pro":{"provider":"databricks","base_model":"databricks-gpt-5-5-pro","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","max_input_tokens":1050000,"max_tokens":1050000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-5":{"provider":"databricks","base_model":"databricks-gpt-5-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1050000,"max_tokens":1050000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-4":{"provider":"databricks","base_model":"databricks-gpt-5-4","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1050000,"max_tokens":1050000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-4-mini":{"provider":"databricks","base_model":"databricks-gpt-5-4-mini","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-4-nano":{"provider":"databricks","base_model":"databricks-gpt-5-4-nano","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-3-codex":{"provider":"databricks","base_model":"databricks-gpt-5-3-codex","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-2":{"provider":"databricks","base_model":"databricks-gpt-5-2","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-1":{"provider":"databricks","base_model":"databricks-gpt-5-1","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5":{"provider":"databricks","base_model":"databricks-gpt-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-mini":{"provider":"databricks","base_model":"databricks-gpt-5-mini","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gpt-5-nano":{"provider":"databricks","base_model":"databricks-gpt-5-nano","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":272000,"max_tokens":272000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.gemini-3-8-flash":{"provider":"databricks","base_model":"databricks-gemini-3-8-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-7-flash":{"provider":"databricks","base_model":"databricks-gemini-3-7-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-6-flash":{"provider":"databricks","base_model":"databricks-gemini-3-6-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true,"deprecation_date":"2026-11-19"},"databricks/system.ai.gemini-3-5-flash":{"provider":"databricks","base_model":"databricks-gemini-3-5-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-5-flash-lite":{"provider":"databricks","base_model":"databricks-gemini-3-5-flash-lite","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-1-pro":{"provider":"databricks","base_model":"databricks-gemini-3-1-pro","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":1000000,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-1-flash-image":{"provider":"databricks","base_model":"databricks-gemini-3-1-flash-image","mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","max_input_tokens":131072,"max_tokens":131072,"max_output_tokens":32768,"supports_vision":true},"databricks/system.ai.gemini-3-1-flash-lite":{"provider":"databricks","base_model":"databricks-gemini-3-1-flash-lite","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":65536,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.gemini-3-pro-image":{"provider":"databricks","base_model":"databricks-gemini-3-pro-image","mode":"image_generation","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","max_input_tokens":65536,"max_tokens":65536,"max_output_tokens":32768,"supports_vision":true},"databricks/system.ai.gemini-3-flash":{"provider":"databricks","base_model":"databricks-gemini-3-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":8192,"supports_vision":true,"supports_audio_input":true},"databricks/system.ai.openjev-qwen35-4b":{"provider":"databricks","base_model":"databricks-openjev-qwen35-4b","mode":"decisions","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_decisions":true},"databricks/system.ai.qwen35-122b-a10b":{"provider":"databricks","base_model":"databricks-qwen35-122b-a10b","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":262144,"max_tokens":262144,"max_output_tokens":25000},"databricks/system.ai.qwen3-next-80b-a3b-instruct":{"provider":"databricks","base_model":"databricks-qwen3-next-80b-a3b-instruct","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":262144,"max_tokens":262144,"max_output_tokens":65536},"databricks/system.ai.grok-4-7":{"provider":"databricks","base_model":"databricks-grok-4-7","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":500000,"max_tokens":500000,"max_output_tokens":500000},"databricks/system.ai.grok-4-6":{"provider":"databricks","base_model":"databricks-grok-4-6","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":500000,"max_tokens":500000,"max_output_tokens":500000},"databricks/system.ai.glm-5-3":{"provider":"databricks","base_model":"databricks-glm-5-3","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072},"databricks/system.ai.glm-5-3-flash":{"provider":"databricks","base_model":"databricks-glm-5-3-flash","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072,"supports_vision":true},"databricks/system.ai.kimi-k3-1-pro":{"provider":"databricks","base_model":"databricks-kimi-k3-1-pro","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1048576,"max_tokens":1048576,"max_output_tokens":131072,"supports_vision":true},"databricks/system.ai.gpt-oss-120b":{"provider":"databricks","base_model":"databricks-gpt-oss-120b","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":131072,"max_tokens":131072,"max_output_tokens":65536},"databricks/system.ai.gpt-oss-20b":{"provider":"databricks","base_model":"databricks-gpt-oss-20b","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":131072,"max_tokens":131072,"max_output_tokens":65536},"databricks/system.ai.gemma-3-12b":{"provider":"databricks","base_model":"databricks-gemma-3-12b","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":131072,"max_tokens":131072,"supports_vision":true},"databricks/system.ai.llama-4-maverick":{"provider":"databricks","base_model":"databricks-llama-4-maverick","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving","max_input_tokens":524288,"max_tokens":32768,"max_output_tokens":32768,"supports_vision":true},"databricks/system.ai.meta-llama-3-3-70b-instruct":{"provider":"databricks","base_model":"databricks-meta-llama-3-3-70b-instruct","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"databricks/system.ai.meta-llama-3-1-8b-instruct":{"provider":"databricks","base_model":"databricks-meta-llama-3-1-8b-instruct","mode":"chat","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"databricks/system.ai.claude-haiku-5-5":{"provider":"databricks","base_model":"claude-haiku-5-5","mode":"chat","source":"https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models#-anthropic-claude-haiku-55","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true,"supports_native_effort":true,"supports_safeguards":true,"knowledge_cutoff":"2026-06-30","tool_use_system_prompt_tokens":286,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_image_input":true,"supports_function_calling":true,"supports_tool_choice":true,"supports_system_messages":true,"supports_sampling_params":false,"supports_reasoning":true,"supports_adaptive_thinking":true,"supports_reasoning_disable":true,"supports_reasoning_effort":false,"thinking_always_on":false,"reasoning_effort_levels":["low","medium","high"],"default_reasoning_effort":"medium","supports_xhigh_reasoning_effort":false,"supports_max_reasoning_effort":false,"metadata":{"notes":"Endpoint databricks-claude-haiku-5-5. Capabilities from the Databricks supported-models page; context window and max output follow Anthropic's published limits. Pricing is Anthropic's first-party list price (https://platform.claude.com/docs/en/about-claude/pricing), tiered at 100k prompt tokens. Databricks had not published DBU rates for Haiku 5.5 as of 2026-10-08, so Databricks billing may differ."}},"databricks/system.ai.claude-haiku-4-5":{"provider":"databricks","base_model":"databricks-claude-haiku-4-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":200000,"max_tokens":200000,"max_output_tokens":64000,"supports_vision":true},"databricks/system.ai.claude-sonnet-5-5":{"provider":"databricks","base_model":"databricks-claude-sonnet-5-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000},"databricks/system.ai.claude-sonnet-5":{"provider":"databricks","base_model":"databricks-claude-sonnet-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-sonnet-4-6":{"provider":"databricks","base_model":"databricks-claude-sonnet-4-6","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":64000,"max_output_tokens":64000,"supports_vision":true},"databricks/system.ai.claude-sonnet-4-5":{"provider":"databricks","base_model":"databricks-claude-sonnet-4-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":200000,"max_tokens":200000,"max_output_tokens":64000,"supports_vision":true},"databricks/system.ai.claude-fable-5-1":{"provider":"databricks","base_model":"databricks-claude-fable-5-1","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-fable-5":{"provider":"databricks","base_model":"databricks-claude-fable-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000},"databricks/system.ai.claude-opus-5-5":{"provider":"databricks","base_model":"databricks-claude-opus-5-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-opus-5":{"provider":"databricks","base_model":"databricks-claude-opus-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-opus-4-8":{"provider":"databricks","base_model":"databricks-claude-opus-4-8","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-opus-4-7":{"provider":"databricks","base_model":"databricks-claude-opus-4-7","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":1000000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-opus-4-6":{"provider":"databricks","base_model":"databricks-claude-opus-4-6","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":1000000,"max_tokens":128000,"max_output_tokens":128000,"supports_vision":true},"databricks/system.ai.claude-opus-4-5":{"provider":"databricks","base_model":"databricks-claude-opus-4-5","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":200000,"max_tokens":200000,"max_output_tokens":64000,"supports_vision":true},"databricks/system.ai.claude-opus-4-1":{"provider":"databricks","base_model":"databricks-claude-opus-4-1","mode":"chat","source":"https://www.databricks.com/product/pricing/proprietary-foundation-model-serving","supports_prompt_caching":true,"max_input_tokens":200000,"max_tokens":200000,"max_output_tokens":32000,"supports_vision":true},"databricks/system.ai.bge-large-en":{"provider":"databricks","base_model":"databricks-bge-large-en","mode":"embedding","source":"https://www.databricks.com/product/pricing/foundation-model-serving"},"bedrock_mantle/in.anthropic.claude-haiku-4-5-20251001-v1:0":{"base_model":"claude-haiku-4-5","mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"knowledge_cutoff":"2025-02-28","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_tool_choice":true,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"]},"bedrock_mantle/in.anthropic.claude-sonnet-5":{"base_model":"claude-sonnet-5","mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":false,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-01-31","provider":"bedrock_mantle","supports_audio_input":false,"supports_audio_output":false,"supports_computer_use":true,"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"]},"cloudflare/workers-ai/@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"base_model":"cloudflare/workers-ai/","mode":"chat","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"cloudflare"},"workers-ai/workers-ai/@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"base_model":"cloudflare/workers-ai/","mode":"chat","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"workers-ai"},"cloudflare/workers-ai/@cf/thinkingmachines/inkling-2-small":{"base_model":"cloudflare/workers-ai/","mode":"chat","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","minimal","low","medium","high","max"],"provider":"cloudflare"},"workers-ai/workers-ai/@cf/thinkingmachines/inkling-2-small":{"base_model":"cloudflare/workers-ai/","mode":"chat","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","minimal","low","medium","high","max"],"provider":"workers-ai"},"azure/cohere-embed-v-4-0":{"base_model":"embed-v-4-0","mode":"embedding","source":"https://models.dev/api.json","max_input_tokens":128000,"max_output_tokens":1536,"max_tokens":1536,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":false,"supports_reasoning":false,"provider":"azure"},"azure/phi-4-reasoning-plus":{"base_model":"phi-4-reasoning-plus","mode":"chat","source":"https://models.dev/api.json","max_input_tokens":32000,"max_output_tokens":4096,"max_tokens":4096,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":false,"supports_reasoning":true,"provider":"azure"},"azure/claude-mythos-5":{"base_model":"claude-mythos-5","mode":"chat","source":"https://models.dev/api.json","supports_prompt_caching":true,"max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"supported_modalities":["text","image","pdf"],"supported_output_modalities":["text"],"supports_vision":true,"supports_pdf_input":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"knowledge_cutoff":"2026-01-31","provider":"azure","supports_safeguards":true},"replicate/topazlabs/image-upscale":{"mode":"image_generation","provider":"replicate","base_model":"image-upscale"},"replicate/prunaai/p-image-upscale":{"mode":"image_generation","provider":"replicate","base_model":"prunaai/p-image-upscale"},"replicate/bytedance/seedance-2.0":{"mode":"video_generation","provider":"replicate","base_model":"seedance-2.0"},"replicate/kwaivgi/kling-v3-video":{"mode":"video_generation","provider":"replicate","base_model":"kling-v3-video"},"replicate/minimax/hailuo-2.3":{"mode":"video_generation","provider":"replicate","base_model":"hailuo-2.3"},"replicate/pixverse/pixverse-v6":{"mode":"video_generation","provider":"replicate","base_model":"pixverse-v6"},"replicate/tencent/hunyuan-3d-3.1":{"mode":"3d","provider":"replicate","base_model":"hunyuan-3d-3.1"},"deepseek/deepseek-v4-flash-0731":{"max_input_tokens":1000000,"max_output_tokens":393216,"server_side_model":"deepseek-flash","max_tokens":393216,"mode":"chat","source":"https://api-docs.deepseek.com/quick_start/pricing","supported_endpoints":["/v1/chat/completions"],"supports_assistant_prefill":true,"supports_function_calling":true,"supports_native_streaming":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"provider":"deepseek","peak_hours":{"timezone":"UTC","windows":[{"days":[1,2,3,4,5],"start":"01:00","end":"04:00"},{"days":[1,2,3,4,5],"start":"06:00","end":"10:00"}]},"base_model":"deepseek-v4-flash","supports_native_effort":true},"azure/deepseek-v4.1-flash-2026-09-10":{"max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"deprecation_date":"2026-12-15","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/deepseek-v4-1-flash-is-coming-to-microsoft-foundry/4556431","provider":"azure","base_model":"deepseek-v4.1-flash"},"azure/DeepSeek-V4.1-Flash-2026-09-10":{"max_input_tokens":1000000,"max_output_tokens":384000,"max_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"deprecation_date":"2026-12-15","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/deepseek-v4-1-flash-is-coming-to-microsoft-foundry/4556431","provider":"azure","base_model":"deepseek-v4.1-flash"},"azure/DeepSeek-V4.1-Flash":{"max_input_tokens":1000000,"max_output_tokens":1048576,"max_tokens":384000,"mode":"chat","supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"deprecation_date":"2026-12-15","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/deepseek-v4-1-flash-is-coming-to-microsoft-foundry/4556431","provider":"azure","base_model":"deepseek-v4.1-flash"},"azure/DeepSeek-V3.2":{"max_input_tokens":128000,"max_output_tokens":131072,"max_tokens":128000,"mode":"chat","source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"deepseek"},"azure/DeepSeek-V3.2-Speciale":{"max_input_tokens":128000,"max_output_tokens":128000,"max_tokens":128000,"mode":"chat","source":"https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure","supports_assistant_prefill":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_tool_choice":true,"provider":"azure","base_model":"deepseek-v3.2"},"ollama/tev1":{"base_model":"tev1","provider":"ollama","model":"tev1:4b","mode":"decisions","supports_decisions":true,"max_input_tokens":2000,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/tev1"},"ollama/tev1:0.8b":{"base_model":"tev1","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":2000,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/tev1:0.8b"},"ollama/nimble":{"base_model":"nimble","provider":"ollama","model":"nimble:9b","mode":"decisions","supports_decisions":true,"max_input_tokens":8192,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/nimble"},"ollama/clef-flash":{"base_model":"clef-flash","provider":"ollama","model":"clef-flash:9b","mode":"decisions","supports_decisions":true,"max_input_tokens":65536,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text","image"],"supports_vision":true,"supports_image_input":true,"supports_multimodal":true,"source":"https://ollama.com/library/clef-flash"},"ollama/clef":{"base_model":"clef","provider":"ollama","model":"clef:27b","mode":"decisions","supports_decisions":true,"max_input_tokens":65536,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text","image"],"supports_vision":true,"supports_image_input":true,"supports_multimodal":true,"source":"https://ollama.com/library/clef"},"ollama/laya":{"base_model":"laya","provider":"ollama","model":"laya:421m-english-mlx-fp16","mode":"decisions","supports_decisions":true,"max_input_tokens":512,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/laya"},"ollama/laya:421m-typed-decisions-mlx-fp16":{"base_model":"laya","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":1024,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/laya:421m-typed-decisions-mlx-fp16"},"ollama/laya:322m-multilingual-mlx-fp16":{"base_model":"laya","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":1024,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/laya:322m-multilingual-mlx-fp16"},"ollama/kev":{"base_model":"kev","provider":"ollama","model":"kev:9b-nvfp4","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev"},"ollama/kev:9b-mlx-bf16":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:9b-mlx-bf16"},"ollama/kev:27b-nvfp4":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:27b-nvfp4"},"ollama/kev:27b-mlx-bf16":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:27b-mlx-bf16"},"ollama/kev:4b-mlx-bf16":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:4b-mlx-bf16"},"ollama/kev:0.8b-mxfp8":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:0.8b-mxfp8"},"ollama/kev:0.8b-mlx-bf16":{"base_model":"kev","provider":"ollama","mode":"decisions","supports_decisions":true,"max_input_tokens":73728,"supported_endpoints":["/v1/systemone"],"supported_modalities":["text"],"supports_vision":false,"source":"https://ollama.com/library/kev:0.8b-mlx-bf16"},"cloudflare/@cf/zai-org/glm-5.3":{"base_model":"glm-5.3","mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/glm-5.3/","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"cloudflare"},"cloudflare/@cf/zai-org/glm-5.3-flash":{"base_model":"glm-5.3-flash","mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/glm-5.3-flash/","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","high","max"],"provider":"cloudflare"},"cloudflare/@cf/deepseek-ai/deepseek-v4-flash-0731":{"base_model":"deepseek-v4-flash","mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/deepseek-v4-flash-0731/","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"cloudflare"},"cloudflare/@cf/deepseek-ai/deepseek-v4-pro-0813":{"base_model":"deepseek-v4-pro","mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/deepseek-v4-pro-0813/","supports_prompt_caching":true,"max_input_tokens":1048576,"max_output_tokens":1048576,"max_tokens":1048576,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["none","low","high","max"],"provider":"cloudflare"},"cloudflare/@cf/qwen/qwen3.8-27b":{"base_model":"qwen3.8-27b","mode":"chat","source":"https://developers.cloudflare.com/workers-ai/models/qwen3.8-27b/","supports_prompt_caching":true,"max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_function_calling":true,"supports_response_schema":true,"supports_reasoning":true,"reasoning_effort_levels":["low","medium","xhigh"],"provider":"cloudflare"},"xiaomi/mimo-v2.6-pro":{"provider":"xiaomi","base_model":"mimo-v2.6-pro","mode":"chat","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go","web_search_billing_unit":"per_request","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_endpoints":["/v1/chat/completions"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true},"xiaomi/mimo-v2.6-flash":{"provider":"xiaomi","base_model":"mimo-v2.6-flash","mode":"chat","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go","web_search_billing_unit":"per_request","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_endpoints":["/v1/chat/completions"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true},"xiaomi/mimo-v2.6-pro-ultraspeed":{"provider":"xiaomi","base_model":"mimo-v2.6-pro-ultraspeed","mode":"chat","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go","web_search_billing_unit":"per_request","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"supported_endpoints":["/v1/chat/completions"],"supports_audio_input":true,"supports_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_video_input":true,"supports_vision":true,"supports_web_search":true},"xiaomi/mimo-v2.5-asr":{"provider":"xiaomi","base_model":"mimo-v2.5-asr","mode":"audio_transcription","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go"},"xiaomi/mimo-v2.5-tts":{"provider":"xiaomi","base_model":"mimo-v2.5-tts","mode":"audio_speech","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go"},"xiaomi/mimo-v2.5-tts-voiceclone":{"provider":"xiaomi","base_model":"mimo-v2.5-tts","mode":"audio_speech","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go"},"xiaomi/mimo-v2.5-tts-voicedesign":{"provider":"xiaomi","base_model":"mimo-v2.5-tts","mode":"audio_speech","source":"https://mimo.mi.com/docs/en-US/price/pay-as-you-go"},"bedrock/claude-haiku-4-5":{"supports_tool_search":true,"max_input_tokens":200000,"max_output_tokens":64000,"max_tokens":64000,"mode":"chat","source":"https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock","supports_assistant_prefill":true,"supports_computer_use":true,"supports_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":true,"supports_native_structured_output":true,"supports_parallel_tool_use_config":true,"prompt_cache_min_tokens":4096,"provider":"bedrock","base_model":"claude-haiku-4-5","supports_web_search":true},"byteplus/dola-seed-2-1-turbo-260628":{"base_model":"dola-seed-2-1-turbo","provider":"byteplus","mode":"chat","max_input_tokens":262144,"max_output_tokens":262144,"max_tokens":262144,"rpm":500,"tpm":1000000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seed-2-0-lite-260428":{"base_model":"seed-2-0-lite","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supported_modalities":["text","image","audio"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seed-2-0-mini-260428":{"base_model":"seed-2-0-mini","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supported_modalities":["text","image","audio"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":true,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seed-2-0-pro-260328":{"base_model":"seed-2-0-pro","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":false,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seed-2-0-lite-260228":{"base_model":"seed-2-0-lite","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/seed-2-0-mini-260215":{"base_model":"seed-2-0-mini","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/seed-2-0-code-preview-260328":{"base_model":"seed-2-0-code-preview","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":131072,"max_tokens":131072,"rpm":30000,"tpm":1500000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":false,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/glm-5-3-flash-260828":{"base_model":"glm-5-3-flash","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"rpm":500,"tpm":1000000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":false,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/glm-5-2-260617":{"base_model":"glm-5-2","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":131072,"max_tokens":131072,"rpm":500,"tpm":1000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-1-flash-260910":{"base_model":"deepseek-v4-1-flash","provider":"byteplus","mode":"chat","peak_hours":{"timezone":"Asia/Shanghai","windows":[{"days":[1,2,3,4,5],"start":"09:00","end":"12:00"},{"days":[1,2,3,4,5],"start":"14:00","end":"18:00"}]},"max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":3000,"tpm":5000000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-pro-ga-260813":{"base_model":"deepseek-v4-pro","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":500,"tpm":1000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-flash-ga-260731":{"base_model":"deepseek-v4-flash","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":500,"tpm":1000000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-pro-260425":{"base_model":"deepseek-v4-pro","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":15000,"tpm":1500000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_response_schema":false,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-flash-260425":{"base_model":"deepseek-v4-flash","provider":"byteplus","mode":"chat","max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":15000,"tpm":1500000,"supported_modalities":["text"],"supported_output_modalities":["text"],"supports_vision":false,"supports_audio_input":false,"supports_response_schema":false,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seed-1-8-251228":{"base_model":"seed-1-8","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":65536,"max_tokens":65536,"rpm":30000,"tpm":1500000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/glm-4-7-251222":{"base_model":"glm-4-7","provider":"byteplus","mode":"chat","max_input_tokens":204800,"max_output_tokens":131072,"max_tokens":131072,"rpm":15000,"tpm":1500000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/deepseek-v3-2-251201":{"base_model":"deepseek-v3-2","provider":"byteplus","mode":"chat","tiered_pricing":[{"range":[0,32000],"input_cost_per_token":2.8e-7,"output_cost_per_token":4.2e-7},{"range":[32001,128000],"input_cost_per_token":5.6e-7,"output_cost_per_token":8.4e-7}],"max_input_tokens":131072,"max_output_tokens":65536,"max_tokens":65536,"rpm":15000,"tpm":1500000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/gpt-oss-120b-250805":{"base_model":"gpt-oss-120b","provider":"byteplus","model":"gpt-oss-120b-250805","mode":"chat","max_input_tokens":98304,"max_output_tokens":65536,"max_tokens":65536,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text"],"supported_output_modalities":["text"],"supported_regions":["ap-southeast-1"],"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_function_calling":true,"supports_system_messages":true,"supports_native_streaming":true,"supports_vision":false,"supports_image_input":false,"supports_video_input":false,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":false,"supports_native_structured_output":false,"supports_prompt_caching":false,"rpm":15000,"tpm":800000,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing","is_deprecated":true},"byteplus/seed-1-6-250915":{"base_model":"seed-1-6","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":65536,"max_tokens":65536,"rpm":15000,"tpm":800000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/seed-1-6-250615":{"base_model":"seed-1-6","provider":"byteplus","model":"seed-1-6-250615","mode":"chat","max_input_tokens":229376,"max_output_tokens":32768,"max_tokens":32768,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supported_regions":["ap-southeast-1"],"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_function_calling":true,"supports_system_messages":true,"supports_native_streaming":true,"supports_vision":true,"supports_image_input":true,"supports_video_input":true,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supports_native_structured_output":true,"supports_prompt_caching":true,"rpm":15000,"tpm":800000,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing","tiered_pricing":[{"range":[0,131072],"input_cost_per_token":2.5e-7,"output_cost_per_token":0.000002,"cache_read_input_token_cost":5e-8},{"range":[131073,262144],"input_cost_per_token":5e-7,"output_cost_per_token":0.000004,"cache_read_input_token_cost":5e-8}],"is_deprecated":true},"byteplus/seed-1-6-flash-250715":{"base_model":"seed-1-6-flash","provider":"byteplus","mode":"chat","max_input_tokens":229376,"max_output_tokens":65536,"max_tokens":65536,"rpm":15000,"tpm":800000,"supports_prompt_caching":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/seed-1-6-flash-250615":{"base_model":"seed-1-6-flash","provider":"byteplus","model":"seed-1-6-flash-250615","mode":"chat","max_input_tokens":229376,"max_output_tokens":32768,"max_tokens":32768,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image","video"],"supported_output_modalities":["text"],"supported_regions":["ap-southeast-1"],"supports_reasoning":true,"supports_reasoning_with_tool_calls":true,"supports_function_calling":true,"supports_system_messages":true,"supports_native_streaming":true,"supports_vision":true,"supports_image_input":true,"supports_video_input":true,"supports_audio_input":false,"supports_audio_output":false,"supports_response_schema":true,"supports_native_structured_output":true,"supports_prompt_caching":true,"rpm":15000,"tpm":800000,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing","tiered_pricing":[{"range":[0,131072],"input_cost_per_token":7.5e-8,"output_cost_per_token":3e-7,"cache_read_input_token_cost":1.5e-8},{"range":[131073,262144],"input_cost_per_token":1e-7,"output_cost_per_token":8e-7,"cache_read_input_token_cost":1.5e-8}],"is_deprecated":true},"byteplus/dreamina-seedance-2-5-260628":{"base_model":"dreamina-seedance-2-5","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"supports_image_input":true,"supports_video_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/dreamina-seedance-2-0-260128":{"base_model":"dreamina-seedance-2-0","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"supports_image_input":true,"supports_video_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/dreamina-seedance-2-0-fast-260128":{"base_model":"dreamina-seedance-2-0-fast","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"supports_image_input":true,"supports_video_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/dreamina-seedance-2-0-mini-260615":{"base_model":"dreamina-seedance-2-0-mini","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image","video"],"supported_output_modalities":["video"],"supports_image_input":true,"supports_video_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedance-1-0-pro-250528":{"base_model":"seedance-1-0-pro","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedance-1-0-pro-fast-251015":{"base_model":"seedance-1-0-pro-fast","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedance-1-5-pro-251215":{"base_model":"seedance-1-5-pro","provider":"byteplus","mode":"video_generation","supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"byteplus/dola-seedream-5-0-pro-260628":{"base_model":"dola-seedream-5-0-pro","provider":"byteplus","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/dola-seedream-5-0-flash-260915":{"base_model":"dola-seedream-5-0-flash","provider":"byteplus","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedream-5-0-lite-260128":{"base_model":"seedream-5-0-lite","provider":"byteplus","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedream-5-0-260128":{"base_model":"seedream-5-0","provider":"byteplus","model":"seedream-5-0-260128","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_size":true,"supported_endpoints":["/v1/images/generations"],"supported_regions":["ap-southeast-1"],"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing"},"byteplus/seedream-4-5-251128":{"base_model":"seedream-4-5","provider":"byteplus","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seedream-4-0-250828":{"base_model":"seedream-4-0","provider":"byteplus","mode":"image_generation","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/seededit-3-0-i2i-250628":{"base_model":"seededit-3-0-i2i","provider":"byteplus","mode":"image_edit","supported_modalities":["text","image"],"supported_output_modalities":["image"],"supports_image_input":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/hyper3d-gen2-260112":{"base_model":"hyper3d-gen2","provider":"byteplus","mode":"3d","supported_modalities":["text","image"],"supports_image_input":true,"rpm":60,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/hitem3d-2-0-251223":{"base_model":"hitem3d-2-0","provider":"byteplus","mode":"3d","supported_modalities":["image"],"supports_image_input":true,"rpm":600,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/skylark-embedding-vision-251215":{"base_model":"skylark-embedding-vision","provider":"byteplus","mode":"embedding","max_input_tokens":131072,"output_vector_size":2048,"supported_modalities":["text","image","video"],"supports_embedding_image_input":true,"rpm":1200,"tpm":1200000,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/skylark-embedding-vision-250615":{"base_model":"skylark-embedding-vision","provider":"byteplus","mode":"embedding","supports_embedding_image_input":true,"deprecation_date":"2026-11-11","source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md","is_deprecated":true},"databricks/system.ai.typesafeai-jev-1-14-0":{"provider":"databricks","base_model":"databricks-typesafeai-jev-1-14-0","mode":"decisions","source":"https://www.databricks.com/product/pricing/foundation-model-serving","supports_decisions":true},"minimax/image-01":{"base_model":"image-01","provider":"minimax","mode":"image_generation","supported_endpoints":["/v1/image_generation"],"supported_modalities":["text","image"],"supported_output_modalities":["image"],"supported_resolutions":["1024x1024","1280x720","1152x864","1248x832","832x1248","864x1152","720x1280","1344x576"],"supports_image_input":true,"supports_image_size":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/image"},"minimax/MiniMax-H3-Regeneration":{"base_model":"MiniMax-H3","provider":"minimax","model":"MiniMax-H3","mode":"video_generation","supported_endpoints":["/v2/video_regeneration"],"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["video"],"supports_image_input":true,"supports_video_input":true,"supports_audio_input":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/video"},"minimax/MiniMax-H3-Context-IR":{"base_model":"MiniMax-H3","provider":"minimax","model":"MiniMax-H3","supported_endpoints":["/v2/h3_context_ir"],"supported_modalities":["text","image","video","audio"],"supported_output_modalities":["text"],"supports_image_input":true,"supports_video_input":true,"supports_audio_input":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/video"},"minimax/MiniMax-Hailuo-2.3":{"base_model":"MiniMax-Hailuo-2.3","provider":"minimax","mode":"video_generation","supported_endpoints":["/v1/video_generation"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/video"},"minimax/MiniMax-Hailuo-2.3-Fast":{"base_model":"MiniMax-Hailuo-2.3-Fast","provider":"minimax","mode":"video_generation","supported_endpoints":["/v1/video_generation"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/video"},"minimax/MiniMax-Hailuo-02":{"base_model":"MiniMax-Hailuo-02","provider":"minimax","mode":"video_generation","supported_endpoints":["/v1/video_generation"],"supported_modalities":["text","image"],"supported_output_modalities":["video"],"supports_image_input":true,"source":"https://platform.minimax.io/docs/pricing/overview#models/video"},"byteplus/deepseek-v4-1-flash-economy-261017":{"base_model":"deepseek-v4-1-flash","provider":"byteplus","mode":"chat","peak_hours":{"timezone":"Asia/Shanghai","windows":[{"days":[1,2,3,4,5],"start":"09:00","end":"12:00"},{"days":[1,2,3,4,5],"start":"14:00","end":"18:00"}]},"max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":3000,"tpm":5000000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"byteplus/deepseek-v4-1-flash-261017":{"base_model":"deepseek-v4-1-flash","provider":"byteplus","mode":"chat","peak_hours":{"timezone":"Asia/Shanghai","windows":[{"days":[1,2,3,4,5],"start":"09:00","end":"12:00"},{"days":[1,2,3,4,5],"start":"14:00","end":"18:00"}]},"max_input_tokens":1048576,"max_output_tokens":393216,"max_tokens":393216,"rpm":3000,"tpm":5000000,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_vision":true,"supports_audio_input":false,"supports_response_schema":true,"supports_function_calling":true,"supports_reasoning":true,"supports_prompt_caching":true,"source":"https://ai.byteplus.com/ark/region:ap-southeast-1/docs/model-pricing.md"},"azure/neural":{"mode":"audio_speech","provider":"azure","base_model":"neural"},"azure/DragonLatestNeural":{"mode":"audio_speech","provider":"azure","base_model":"DragonLatestNeural"},"azure/DragonHDFlashLatestNeural":{"mode":"audio_speech","provider":"azure","base_model":"DragonHDFlashLatestNeural"},"azure/DragonHDLatestNeural":{"mode":"audio_speech","provider":"azure","base_model":"DragonHDLatestNeural"},"azure/DragonHDOmniLatestNeural":{"mode":"audio_speech","provider":"azure","base_model":"DragonHDOmniLatestNeural"},"azure/fast-transcription":{"mode":"audio_transcription","provider":"azure","base_model":"fast-transcription"},"azure/llm-speech":{"mode":"audio_transcription","provider":"azure","base_model":"llm-speech"},"azure/short-audio":{"mode":"audio_transcription","provider":"azure","base_model":"short-audio"},"azure/MAI-Transcribe-2":{"mode":"audio_transcription","provider":"azure","base_model":"MAI-Transcribe-2"},"azure/gpt-6.1-terra-2026-10-26":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/us/gpt-6.1-terra-2026-10-26":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["[/v1/chat/completions](https://www.getbifrost.ai/v1/chat/completions)","[/v1/responses](https://www.getbifrost.ai/v1/responses)"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/eu/gpt-6.1-terra-2026-10-26":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["[/v1/chat/completions](https://www.getbifrost.ai/v1/chat/completions)","[/v1/responses](https://www.getbifrost.ai/v1/responses)"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"azure/apac/gpt-6.1-terra-2026-10-26":{"max_input_tokens":922000,"max_output_tokens":128000,"mode":"chat","source":"https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/gpt-6-1-terra-now-generally-available-in-microsoft-foundry-build-agents-that-see/4567769","supported_endpoints":["[/v1/chat/completions](https://www.getbifrost.ai/v1/chat/completions)","[/v1/responses](https://www.getbifrost.ai/v1/responses)"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"provider":"azure","base_model":"gpt-6.1-terra"},"openai/gpt-6-luna":{"mode":"decisions","provider":"openai","base_model":"gpt-6-luna","source":"https://developers.openai.com/api/docs/guides/decisions","supported_endpoints":["/v1/decisions"],"supported_modalities":["text","image"],"supports_vision":true,"supports_decisions":true,"max_input_tokens":1050000,"max_output_tokens":128000,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_modes":["standard","pro"],"regional_processing_uplift_multiplier_eu":1.1,"regional_processing_uplift_multiplier_us":1.1,"supported_output_modalities":["text"],"supports_computer_use":true,"supports_function_calling":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_none_reasoning_effort":true,"supports_parallel_function_calling":true,"supports_pdf_input":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_web_search":true,"supports_xhigh_reasoning_effort":true,"web_search_billing_unit":"per_request","regional_endpoint_uplift_multiplier":1.1,"default_reasoning_effort":"medium","reasoning_effort_levels":["none","low","medium","high","xhigh","max"],"supports_image_input":true,"supports_reasoning_with_tool_calls":false,"supports_low_reasoning_effort":true,"supports_native_structured_output":true,"supports_file_search":true,"supports_service_tier":true,"metadata":{"notes":"For prompts over 272K input tokens, OpenAI charges 2x input/cache and 1.5x output for the full request. Batch and Flex are 50% of Standard. Fast is 2x the applicable rate. The current Bifrost schema has no dedicated Fast-above-272K pricing fields, so the Fast fields here represent short-context Fast pricing."},"supports_reasoning_effort":true,"is_reasoning_model":true,"supports_reasoning_disable":true,"thinking_always_on":false,"always_reasoning":false,"supported_reasoning_contexts":["auto","current_turn","all_turns"],"min_output_tokens":16,"unsupported_fields":{"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"conditionally_unsupported_fields":{"temperature":"when_effort_none","top_p":"when_effort_none","logprobs":"when_effort_none","top_logprobs":"when_effort_none"},"service_tiers":["auto","default","flex","priority","fast"]},"gemini/antigravity-preview-09-2026":{"provider":"gemini","base_model":"antigravity-preview-09-2026","model":"antigravity-preview-09-2026","mode":"responses","max_input_tokens":1048576,"max_output_tokens":65536,"max_tokens":65536,"web_search_billing_unit":"per_query","supports_function_calling":true,"supports_web_search":true,"supports_url_context":true,"supports_code_execution":true,"supports_file_search":false,"supports_computer_use":false,"supports_response_schema":false,"supports_native_structured_output":false,"supports_native_streaming":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_sampling_params":false,"supports_vision":true,"supports_image_input":true,"supports_pdf_input":false,"supports_audio_input":false,"supports_video_input":false},"anthropic/container":{"mode":"chat","source":"https://platform.claude.com/docs/en/about-claude/pricing#code-execution-tool","provider":"anthropic","base_model":"container"},"vertex_ai/chirp-3-hd":{"provider":"vertex_ai","mode":"audio_speech","base_model":"chirp-3-hd"},"elevenlabs/scribe_v2_realtime":{"provider":"elevenlabs","base_model":"scribe-v2-realtime","mode":"audio_transcription","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_native_streaming":true},"elevenlabs/scribe_v3_live":{"provider":"elevenlabs","base_model":"scribe-v3-live","mode":"audio_transcription","source":"https://elevenlabs.io/pricing/api","supported_endpoints":["/v1/realtime"],"supported_modalities":["audio"],"supported_output_modalities":["text"],"supports_audio_input":true,"supports_native_streaming":true},"bedrock_mantle/anthropic.claude-mythos-5-2":{"base_model":"claude-mythos-5-2","provider":"bedrock_mantle","mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"prompt_cache_min_tokens":512,"tool_use_system_prompt_tokens":346,"supports_reasoning":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_native_effort":true,"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_output_config":true,"supports_sampling_params":false,"supports_assistant_prefill":false,"supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_parallel_tool_use_config":true,"supports_computer_use":true,"supports_mid_conversation_system":true,"supports_system_messages":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_vision":true,"supports_image_input":true,"supports_web_search":true,"supports_safeguards":true,"supports_native_streaming":true,"supports_audio_input":false,"supports_audio_output":false,"supported_modalities":["text","image"],"supported_output_modalities":["text"],"token_count_model":{"model":"anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"us-east-1"},"supported_regions":["us-east-1","us-east-2","us-west-1","us-west-2","ca-central-1","ca-west-1","us-gov-west-1","us-gov-east-1","eu-central-1","eu-central-2","eu-north-1","eu-south-1","eu-south-2","eu-west-1","eu-west-2","eu-west-3","ap-east-2","ap-northeast-1","ap-northeast-2","ap-northeast-3","ap-south-1","ap-south-2","ap-southeast-1","ap-southeast-2","ap-southeast-3","ap-southeast-4","ap-southeast-5","ap-southeast-6","ap-southeast-7","il-central-1","me-central-1","me-south-1","af-south-1","sa-east-1","mx-central-1"]},"bedrock_mantle/us-gov-west-1/anthropic.claude-mythos-5-2":{"base_model":"claude-mythos-5-2","provider":"bedrock_mantle","mode":"chat","source":"https://aws.amazon.com/bedrock/pricing/","max_input_tokens":1000000,"max_output_tokens":128000,"max_tokens":128000,"prompt_cache_min_tokens":512,"tool_use_system_prompt_tokens":346,"supports_reasoning":true,"supports_adaptive_thinking":true,"thinking_always_on":true,"supports_reasoning_disable":false,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"high","supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_native_effort":true,"bedrock_converse_supports_strict_tools":false,"bedrock_output_config_effort_ceiling":"xhigh","supports_output_config":true,"supports_sampling_params":false,"supports_assistant_prefill":false,"supports_forced_tool_choice":false,"supports_forced_tool_use":false,"supports_function_calling":true,"supports_tool_choice":true,"supports_parallel_tool_use_config":true,"supports_computer_use":true,"supports_mid_conversation_system":true,"supports_native_structured_output":true,"supports_response_schema":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_vision":true,"supports_safeguards":true,"supports_audio_input":false,"supports_audio_output":false,"token_count_model":{"model":"anthropic.claude-mythos-5-2","litellm_provider":"bedrock_mantle","region":"us-gov-west-1"},"supported_regions":["us-gov-west-1","us-gov-east-1"]},"sarvam/sarvam-105b":{"max_input_tokens":128000,"max_output_tokens":128000,"mode":"chat","provider":"sarvam","base_model":"sarvam-105b","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_audio_input":false,"supports_video_input":false},"sarvam/sarvam-105b-conversations":{"max_input_tokens":32000,"max_output_tokens":32000,"mode":"chat","provider":"sarvam","base_model":"sarvam-105b-conversations","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":false,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_audio_input":false,"supports_video_input":false},"sarvam/glm5.3":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","provider":"sarvam","base_model":"glm5.3","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":false,"supports_audio_input":false,"supports_video_input":false},"sarvam/glm5.3-flash":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","provider":"sarvam","base_model":"glm5.3-flash","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_video_input":false},"sarvam/gemma4":{"max_input_tokens":262144,"max_output_tokens":262144,"mode":"chat","provider":"sarvam","base_model":"gemma4","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_video_input":false},"sarvam/deepseekv4.1-flash-1017":{"max_input_tokens":1048576,"max_output_tokens":1048576,"mode":"chat","provider":"sarvam","base_model":"deepseekv4.1-flash-1017","source":"https://www.sarvam.ai/token-factory/pricing","supports_function_calling":true,"supports_parallel_function_calling":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_response_schema":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_vision":true,"supports_audio_input":false,"supports_video_input":false},"sarvam/sarvam-prism":{"max_input_tokens":32768,"mode":"decisions","provider":"sarvam","base_model":"sarvam-prism","source":"https://docs.sarvam.ai/api/token-factory/models/sarvam-prism","supports_decisions":true,"supports_vision":false},"soniox/tts-rt-v2":{"base_model":"tts-rt-v2","mode":"audio_speech","provider":"soniox","source":"https://soniox.com/pricing","metadata":{"calculation":"$4.00 per 1M input text tokens x about 0.3 tokens per character = $0.0000012 per character; $21.50 per 1M output audio tokens x about 30,000 tokens per hour of speech = $0.000179167 per second of audio","notes":"Soniox bills tokens, which Bifrost does not receive; the per-character and per-second rates use the conversions on soniox.com/pricing"},"supported_endpoints":["/v1/audio/speech"]},"elevenlabs/scribe_v3":{"base_model":"scribe-v3","mode":"audio_transcription","provider":"elevenlabs","source":"https://elevenlabs.io/pricing/api","metadata":{"calculation":"$0.20/hour = $0.0000555556 per second","original_pricing_per_hour":0.2},"supported_endpoints":["/v1/audio/transcriptions"]},"elevenlabs/scribe_v2_medical":{"base_model":"scribe-v2-medical","mode":"audio_transcription","source":"https://elevenlabs.io/pricing/api","metadata":{"calculation":"$0.20/hour = $0.0000555556 per second","original_pricing_per_hour":0.2},"supported_endpoints":["/v1/audio/transcriptions"]},"azure/gpt-6.2-astra-2026-11-17":{"base_model":"gpt-6.2-astra","provider":"azure","mode":"chat","max_input_tokens":922000,"max_output_tokens":128000,"min_output_tokens":16,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_service_tier":false,"supports_native_streaming":true,"supports_system_messages":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_custom_tools":true,"supports_tool_search":true,"supports_computer_use":true,"supports_web_search":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_image_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"reasoning_contexts":["auto","current_turn","all_turns"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"web_search_billing_unit":"per_request","regional_processing_uplift_multiplier_us":1.1,"regional_processing_uplift_multiplier_eu":1.2,"max_tokens":128000,"source":"https://ai.azure.com/catalog/models/gpt-6.2-astra","metadata":{"notes":"PROVISIONAL: Azure has not yet published GPT-6.2 prices (catalog version 2026-11-17 links to https://aka.ms/AzureOAIpricing with no figures). Global Standard is set to OpenAI's Standard rates, as Azure matched OpenAI for GPT-6 Astra/Sol. Data Zone uplift follows the GPT-6 pattern: US 1.1x, EU 1.2x. Over 272K input tokens: 2x input/cache, 1.5x output for the full request. Knowledge cutoff: 2026-05-18. API rejects reasoning effort 'none'. Re-verify once Azure posts pricing."}},"azure/gpt-6.2-sol-2026-11-17":{"base_model":"gpt-6.2-sol","provider":"azure","mode":"chat","max_input_tokens":922000,"max_output_tokens":128000,"min_output_tokens":16,"supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supports_service_tier":false,"supports_native_streaming":true,"supports_system_messages":true,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_custom_tools":true,"supports_tool_search":true,"supports_computer_use":true,"supports_web_search":true,"supports_file_search":true,"supports_code_execution":true,"supports_image_gen_tool":true,"supports_response_schema":true,"supports_native_structured_output":true,"supports_vision":true,"supports_image_input":true,"supports_pdf_input":true,"supports_prompt_caching":true,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_reasoning":true,"supports_reasoning_effort":true,"is_reasoning_model":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"default_reasoning_effort":"medium","supports_none_reasoning_effort":false,"supports_minimal_reasoning_effort":false,"supports_low_reasoning_effort":true,"supports_xhigh_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_reasoning_disable":false,"thinking_always_on":true,"always_reasoning":true,"supports_reasoning_with_tool_calls":false,"reasoning_contexts":["auto","current_turn","all_turns"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_sampling_params":false,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true},"web_search_billing_unit":"per_request","regional_processing_uplift_multiplier_us":1.1,"regional_processing_uplift_multiplier_eu":1.2,"max_tokens":128000,"source":"https://ai.azure.com/catalog/models/gpt-6.2-sol","metadata":{"notes":"PROVISIONAL: Azure has not yet published GPT-6.2 prices (catalog version 2026-11-17 links to https://aka.ms/AzureOAIpricing with no figures). Global Standard is set to OpenAI's Standard rates, as Azure matched OpenAI for GPT-6 Astra/Sol. Data Zone uplift follows the GPT-6 pattern: US 1.1x, EU 1.2x. Over 272K input tokens: 2x input/cache, 1.5x output for the full request. Knowledge cutoff: 2026-06-01. API rejects reasoning effort 'none'. Re-verify once Azure posts pricing."}},"bedrock/global.openai.gpt-6.2-astra":{"base_model":"gpt-6.2-astra","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default","ultrafast"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock/us.openai.gpt-6.2-astra":{"base_model":"gpt-6.2-astra","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default","ultrafast"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock/openai.gpt-6.2-astra":{"base_model":"gpt-6.2-astra","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default","ultrafast"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":true,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock/global.openai.gpt-6.2-sol":{"base_model":"gpt-6.2-sol","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock/us.openai.gpt-6.2-sol":{"base_model":"gpt-6.2-sol","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"bedrock/openai.gpt-6.2-sol":{"base_model":"gpt-6.2-sol","mode":"chat","provider":"bedrock","always_reasoning":true,"default_reasoning_effort":"medium","is_reasoning_model":true,"max_input_tokens":1050000,"max_output_tokens":128000,"max_tokens":128000,"min_output_tokens":16,"prompt_cache_min_tokens":1024,"reasoning_contexts":["auto","current_turn","all_turns"],"reasoning_effort_levels":["low","medium","high","xhigh","max"],"service_tiers":["auto","default"],"source":"https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-openai.html","supported_endpoints":["/v1/chat/completions","/v1/responses"],"supported_modalities":["text","image"],"supported_output_modalities":["text"],"supported_reasoning_contexts":["auto","current_turn","all_turns"],"supports_computer_use":true,"supports_custom_tools":true,"supports_function_calling":true,"supports_image_input":true,"supports_low_reasoning_effort":true,"supports_max_reasoning_effort":true,"supports_minimal_reasoning_effort":false,"supports_native_streaming":true,"supports_native_structured_output":true,"supports_none_reasoning_effort":false,"supports_parallel_function_calling":true,"supports_pdf_input":false,"supports_prompt_cache_breakpoint":true,"supports_prompt_cache_breakpoints":true,"supports_prompt_caching":true,"supports_reasoning":true,"supports_reasoning_disable":false,"supports_reasoning_effort":true,"supports_reasoning_with_tool_calls":false,"supports_response_schema":true,"supports_sampling_params":false,"supports_service_tier":false,"supports_system_messages":true,"supports_tool_choice":true,"supports_tool_search":true,"supports_vision":true,"supports_web_search":false,"supports_xhigh_reasoning_effort":true,"thinking_always_on":true,"unsupported_fields":{"temperature":true,"top_p":true,"logprobs":true,"top_logprobs":true,"stop":true,"presence_penalty":true,"frequency_penalty":true,"logit_bias":true,"web_search_options":true}},"gemini-3.1-flash-live-preview":{"successor_model":"gemini/gemini-3.8-live","max_input_tokens":131072,"max_output_tokens":65536,"mode":"realtime","source":"https://ai.google.dev/gemini-api/docs/pricing","supported_endpoints":["/v1/realtime"],"supported_modalities":["text","image","audio","video"],"supported_output_modalities":["text","audio"],"supports_audio_input":true,"supports_audio_output":true,"supports_function_calling":true,"supports_vision":true,"supports_web_search":true,"tpm":250000,"rpm":10,"gemini_audio_only_live":true,"supports_response_schema":false,"supports_reasoning":true,"provider":"gemini","base_model":"gemini-3.1-flash-live"},"glm-5p3":{"max_input_tokens":1048576,"max_output_tokens":128000,"mode":"chat","source":"https://api.fireworks.ai/v1/serverless/models","supports_function_calling":true,"supports_mid_conversation_system":true,"supports_reasoning":true,"supports_response_schema":true,"supports_tool_choice":true,"supports_vision":false,"provider":"fireworks_ai","base_model":"glm-5.3","supports_prompt_caching":true,"max_tokens":128000,"supports_reasoning_effort":true,"reasoning_effort_levels":["low","medium","high","xhigh","max"],"reasoning_effort_renames":{"none":"low"},"supports_reasoning_disable":false,"model_parameters":[{"id":"max_tokens","label":"Max Tokens","helpText":"The maximum number of tokens that can be generated in the Result.","type":"number","default":4096,"range":{"min":1,"max":128000},"supported_endpoints":["/v1/chat/completions","/v1/completions"]}]},"claude-3-7-sonnet-20250219":{"deprecation_date":"2026-02-19","is_deprecated":true},"claude-3-opus-20240229":{"deprecation_date":"2026-01-05","is_deprecated":true},"claude-opus-4-20250514":{"deprecation_date":"2026-06-15","is_deprecated":true}}